JP7006111B2 - Devices, methods, and programs for locating, devices, methods, and programs for displaying images. - Google Patents

Description

The present invention relates to a device, method, and program for determining a position, a device, method, and program for displaying an image.

Patent Document 1 includes a scanning means for scanning an electron beam and an electron detector for detecting electrons emitted from a sample scanned by the electron beam, and scans an electron image based on the detection result from the electron detector. Disclosed is an apparatus including a scanning electron microscope for obtaining an optical image and an optical microscope for irradiating a sample with illumination light and receiving reflected light from the sample to obtain an optical image.

U.S. Pat. No. 8791415

The apparatus of the first aspect of the technique of the present disclosure comprises a detection unit for detecting at least one sample section in a first image obtained by photographing a substrate on which a plurality of sample sections are arranged, and a predetermined device. A determination unit for determining the position of the detected sample section using a reference position is provided.

In the method of the second aspect of the technique of the present disclosure, the detection unit detects and determines at least one of the sample sections in the first image obtained by photographing the substrate on which a plurality of sample sections are arranged. The unit determines the position of the detected sample section using a predetermined reference position.

The program of the third aspect of the technique of the present disclosure causes the computer to function as the detection unit and the determination unit of the device of the first aspect.

The apparatus of the fourth aspect of the technique of the present disclosure detects a predetermined shape using the first parameter in the first image and locates the selected position in the second image showing the detected result. In a part of the region of the second image including, the detection unit for detecting the predetermined shape using the second parameter and the position of the detected predetermined shape using the predetermined reference position. It is provided with a decision unit for determining.

The apparatus of the fifth aspect of the technique of the present disclosure uses the first parameter in the first image to display a detection unit that detects a predetermined shape and a second image showing the detected result. It has a display unit, and the detection unit detects the predetermined shape using the second parameter in a part of the region of the second image including the position selected in the second image. do.

In the method of the sixth aspect of the technique of the present disclosure, the detection unit detects a predetermined shape using the first parameter in the first image, and the detection unit shows the detected result. The predetermined shape is detected using the second parameter in a part of the area of the second image including the position selected in the second image.

In the method of the seventh aspect of the technique of the present disclosure, the detection unit detects a predetermined shape using the first parameter in the first image, and the detection unit shows the detected result. In a part of the area of the second image including the position selected in the second image, the second parameter is used to detect the predetermined shape, and the determination unit uses the predetermined reference position to detect the predetermined shape. The position of the detected predetermined shape is determined.

The program of the eighth aspect of the technique of the present disclosure causes the computer to function as the detector according to the fourth or fifth aspect.

It is a block diagram of a CLEM (Correlative Light and Electron Microscopy) system 10. It is a figure which showed the structure of the function of the slice detection processing program executed by the CPU 22 of the image processing apparatus 16. It is a figure which shows the wafer 60 in which a plurality of wafer chips 64 are arranged. It is a figure which shows the image 34D of a wafer chip 64. It is a flowchart of a slice detection processing program. It is a flowchart of the slice template creation processing program of step 80 of FIG. It is a figure which shows the order which specifies the corner of a slice template. It is a figure which shows the order which specifies the corner of a slice template. It is a figure which shows the order which specifies the corner of a slice template. It is a figure explaining how the position of the clicked point is rotated, and the coordinate of the clicked point is corrected so that the base of the selected slice 93 becomes horizontal. It is a figure explaining how the position of the clicked point is rotated, and the coordinate of the clicked point is corrected so that the base of the selected slice 93 becomes horizontal. It is a figure explaining how the position of the clicked point is rotated, and the coordinate of the clicked point is corrected so that the base of the selected slice 93 becomes horizontal. It is a flowchart of the auto detection processing program of step 82 of FIG. It is a figure which shows the display screen 34G of a display part 34. It is a figure which shows how the number of edges to be extracted differs according to the designated threshold value. It is a figure which shows how the number of edges to be extracted differs according to the designated threshold value. It is a figure which shows how the number of edges to be extracted differs according to the designated threshold value. Is a diagram showing an image 34D of a wafer chip 64 and a luminance gradient direction table 34DE showing the direction of the luminance gradient for each pixel. Is a diagram showing a state of creating a filter. Is a diagram showing a score table 34DM of values (scores) representing the certainty (similarity) that the pixels of the image 34D of the wafer chip 64 are the pixels of the slice 66. It is a figure which shows the state of obtaining the score in the evaluation area 102 in the image 34D of a wafer chip 64. It is a figure which shows the slice candidate 112. It is a figure which shows the slice 114. It is a figure which shows the slice 114. It is the figure which arranged the slice candidate 116. It is a figure which shows that the slice candidate 116 is not detected as a slice. It is the figure which arranged the slice candidate 118. It is a figure which shows the slice 114 and the slice 120. It is a figure which shows the mode which specifies the order of slice rows 132, 134, and the order of each slice of slice rows 132, 134. It is a figure which shows the coordinates of the point of each corner of each slice 66 based on the X-axis and Y-axis with respect to 3 points 62A, 62B, 62C to which the absolute position marker is attached, which was defined on the wafer 60. It is a figure which shows the coordinates of the point of each corner of each slice 66. It is a flowchart of the semi-manual detection processing program executed after the slice confirmation processing of step 82E of FIG. 7 of 2nd Embodiment. It is a figure which shows the appearance that the point 152P1 of the slice recognized as a leaking slice is clicked. It is a figure which shows the appearance that the square area 154R about the clicked point 152P1 was cut out as a search area. It is a figure which shows the clicked point 152P1. It is a figure which shows the predetermined area 152PR including the clicked point 152P1. It is a figure which shows the histogram of the luminance of the whole image of a wafer chip. It is a figure which shows the edge 158E0 of a slice which was not detected by auto detection. It is a figure which shows the mode that the slice template 92 is moved in the search area 154R, and the template matching is performed. It is a figure which shows the shape of a slice template 92. It is a figure which shows the search area 154R of a square centering on a clicked point. It is a figure which shows the slice 166X1 and 166X2 which were already detected in the auto detection process, and the slice 166N which was detected and confirmed by a semi-manual detection process. It is a flowchart of the auto detection processing program in the 2nd modification. It is a figure which shows the image 34D of a wafer chip 64. It is a figure which shows the area 82G1 which a user moves a filter 98 for a designated matching process through an input part 36. It is a figure which shows the detected slice. It is a figure which shows the range 82GR1 of the image which performs a matching process. It is a figure which shows the point 82GP1 to 82GP3 at the end of the range of the image which performs a matching process. It is a figure which shows the list of the edge extraction result in 250, 200, 150, 100 which are four predetermined parameters.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(First Embodiment)
As shown in FIG. 1A, the CLEM (Correactive Light and Electron Microscope) system 10 includes an optical microscope 12, an electron microscope 14, and an image processing device 16 connected to the optical microscope 12 and the electron microscope 14. .. Each of the optical microscope 12 and the electron microscope 14 may be equipped with a computer, for example, a personal computer. The image processing device 16 is an example of a position determining device.

The optical microscope 12 acquires the data of the two-dimensional structure or the three-dimensional structure of the neuron by expressing the fluorescent dye only in the desired part of the sample, for example, the neuron of the cell and adjusting the focus three-dimensionally.

The electron microscope 14 irradiates the sample with an electron beam and observes the reflected electrons to acquire data at a predetermined position on the surface of the sample. By the way, in the electron microscope 14, since the data on the surface of the sample is acquired, the data on the three-dimensional structure of the sample cannot be acquired. Therefore, in order to acquire the data of the three-dimensional structure of the sample with the electron microscope 14, a microtome (not shown) is used. Specifically, the electron-stained sample is hardened with a resin, the upper, lower, left and right ends are scraped off so that the cross section is trapezoidal, and sliced by a microtome to make a plurality of thin trapezoidal slices. Multiple slices are placed side by side on a wafer chip. The electron microscope 14 acquires data at a predetermined position on the surface of each slice, and constructs data of a three-dimensional structure based on the data on the surface of each slice.
The slice is an example of the "sample section" of the technique of the present disclosure.

Then, the image processing apparatus 16 displays the data of the three-dimensional structure of the sample constructed by the electron microscope 14 and the data of the three-dimensional structure of the sample obtained by the optical microscope 12. The image processing apparatus 16 may display the data of the three-dimensional structure of the sample constructed by the electron microscope 14 in combination with the data of the three-dimensional structure of the sample obtained by the optical microscope 12. For example, the entire sample may be displayed, and only a desired portion of the sample may be displayed in different colors.

Next, the configuration of the electrical system of the image processing device 16 will be described. The image processing device 16 includes a computer 20. The computer 20 includes a CPU 22, a ROM 24, a RAM 26, and an input / output (I / O) port 28. The CPU 22 to the input / output port 28 are connected to each other by a bus 30. A secondary storage device 32, a display unit 34, and an input unit 36 are connected to the input / output port 28. Further, the optical microscope 12 and the electron microscope 14 are connected to the input / output port 28 via the communication interface (I / F) 38. The slice detection processing program described later is stored in the ROM 24 or the secondary storage device 32.
The display unit 34 is an example of the display unit of the technique of the present disclosure.

Next, with reference to FIG. 1B, the configuration of the function of the slice detection processing program executed by the CPU 22 of the image processing apparatus 16 will be described. As shown in FIG. 1B, the slice detection processing program includes a creation function, a detection function, an adjustment function, an ordering function, a designation function, a determination function, and a transmission function. When the CPU 22 of the image processing device 16 executes each function, the CPU 22 of the image processing device 16 has a creation unit 42, a detection unit 44, an adjustment unit 46, an ordering unit 48, a designation unit 50, a determination unit 52, and a transmission unit. Functions as 54.
The detection unit 44 is an example of the detection unit of the technology of the present disclosure, and the determination unit 52 is an example of the determination unit of the technology of the present disclosure.

Next, with reference to FIG. 2, a wafer 60 in which a plurality of wafer chips 64 are arranged will be described. As shown in FIG. 2A, a plurality of wafer chips 64 are arranged on the wafer 60. As shown in FIG. 2B, a plurality of slices 66 are arranged on the wafer chip 64. As described above, the electron-stained sample is hardened with a resin, the upper, lower, left and right edges are scraped off so that the cross section is trapezoidal, and the sample is sliced by a microtome to form a plurality of slices 66, and the plurality of slices 66 are formed. Arrange them side by side on the wafer chip 64. The electronic staining may be performed after the sample is hardened with a resin.

The shape of the slice 66 is trapezoidal in order to add a cut to indicate the direction. Some of the slices 66 are not exactly trapezoidal and are distorted or thin (or thicker) when sliced.

The sample is continuously sliced by a microtome to form multiple slices 66. When the sample is sliced continuously, the slices already formed are repeatedly extruded onto the water by the slices formed next, so that the plurality of slices 66 are often arranged side by side. A plurality of slices 66 arranged side by side on water are scooped up and arranged side by side on the wafer chip 64. There is also a slice 93 that exists alone. The orientation in which each slice 66 is arranged may be messy. As shown in FIG. 2B, another plurality of slices 68 may also be arranged.

The wafer chip 64 is photographed with an optical microscope 12 or an optical microscope different from the optical microscope 12. The range that can be photographed with an optical microscope may be narrower than that of the wafer chip 64. In that case, the optical microscope captures the entire wafer chip while shifting the imaging region, joins the captured images, and acquires the image 34D of the wafer chip 64 of FIG. 2B. The magnification may be adjusted with another optical microscope to capture the wafer chip 64 as a single image. The image 34D of the wafer chip 64 may be acquired by the optical microscope 12, the electron microscope 14, or another image acquisition device such as a camera.

As shown in FIG. 2A, on the wafer 60, three points 62A, 62B, and 62C are provided as absolute position markers in order to set a reference position common to the optical microscope 12, the electron microscope 14, and the image processing apparatus 16. It has been granted. When the optical microscope 12, the electron microscope 14, and the image processing apparatus 16 acquire an image of the slice 66 mounted on the wafer 60 or the wafer chip 64, the position of the slice 66 is given an absolute position marker 3 It can be specified with reference to points 62A, 62B, and 62C.

Therefore, when the image processing apparatus 16 specifies the position of the slice 66 with reference to the three points 62A, 62B, 62C to which the absolute position marker is attached, the position of the specified slice 66 is the three points 62A, 62B, 62C. As a reference, the optical microscope 12 and the electron microscope 14 can be grasped in common.

Next, the operation of this embodiment will be described. As shown in FIG. 3, in the slice detection processing program, in step 80, the creation unit 42 creates a slice template. Specifically, when the user specifies the shape of the slice template, the creating unit 42 detects the designation of the shape of the slice template in step 80A of FIG. For example, the image 34D of the wafer chip 64 shown in FIG. 2B is displayed on the display unit 34 of the image processing apparatus 16. The user selects a slice 93 having a shape relatively equal to a trapezoid as a slice template from a plurality of slices 66 in the image 34D of the wafer chip 64, and clicks the corners of the slice 93 in order via the input unit 36. do. In step 80A, the creating unit 42 detects the position of the clicked point and determines the shape of the slice template. The click order of the corners may be counterclockwise from the lower left corner of the slice, for example, as shown in FIGS. 5A and 5B. This is because the line connecting the first clicked point and the second clicked point is the base of the trapezoid. In order to determine the base of the trapezoid, a line connecting the first clicked point and the second clicked point counterclockwise from the lower left corner of the slice may be used. The order of clicking may be arbitrary. In addition, the base may be specified by the user after clicking.
The orientation of the slice does not matter. In the example shown in FIGS. 5A and 5B, the number of corners is four. Due to the formation of orientation cuts, the slices may be pentagonal, in which case there are 5 corners, as shown in FIG. 5C.
The image 34D of the wafer chip 64 is an example of the first image of the technique of the present disclosure.

In step 80B of FIG. 4, the creating unit 42 corrects the coordinates of the clicked point detected in step 80A. Specifically, the position of the clicked point is rotated so that the base of the selected slice 93 shown in FIG. 6A is horizontal as shown in FIG. 6B. As shown in FIG. 6C, the origins of the x and y coordinates are set at the base of the slice 93, and the base is clicked so that the base is along the x-axis and the vertical side is parallel to the y-axis and perpendicular to the base. Correct the coordinates of the point. Specifically, the position on the x-axis of the lower left point 93P1 of the slice 93 and the position on the x-axis of the upper left point 93P4 are made the same. More specifically, assuming that the positions on the x-axis of the lower left point 93P1 and the upper left point 93P4 of the slice 93 before the coordinates are corrected are x1 and x4, respectively, the lower left point 93P1 and the upper left point of the slice 93 are taken. The corrected position of 93P4 is set to (x1 + x4) / 2. Similarly, assuming that the positions on the x-axis of the lower right point 93P2 and the upper right point 93P3 of the slice 93 before the coordinates are corrected are x2 and x3, respectively, the lower right point 93P2 and the upper right point 93P3 of the slice 93 are taken. The corrected position of is set to (x2 + x3) / 2. The origin of the x and y coordinates set at the base of the slice 93 is (x1 + x2 + x3 + x4) / 4. When the coordinate correction process of step 80B in FIG. 4 is completed, the slice detection process proceeds to step 82 of FIG.

In step 82 of FIG. 3, the detection unit 44 executes automatic detection. In the automatic detection, the search position limited user assist process in step 81 may be added as in the modification described later.
Specifically, the auto detection in step 82 of FIG. 3 is in step 82A of FIG. 7, and the detection unit 44 extracts an edge. Hereinafter, the edge extraction process will be described by exemplifying the Canny method, but other methods may be used.

As shown in FIG. 8A, the display screen 34G of the display unit 34 includes an area for displaying the image 34D of the wafer chip 64, a slider movement display area 34SP for moving the slider 34S and setting parameters for edge extraction. Has. The parameter for edge extraction may be, for example, a threshold in the Canny method. The display screen 34G of the display unit 34 further has a set value display area 34K for displaying a set value determined by the position of the slider 34S.
In step 82A, the detection unit 44 detects (extracts) the edge by using, for example, the Canny method. In the Canny method, 1) smoothing processing using a Gaussian filter, 2) calculation of the magnitude of the luminance gradient and the direction of the luminance gradient using a Sobel filter, etc., 3) Non-maximum Threshold processing, 4) Hysteresis Throd processing. However, the parameter for edge extraction is used, for example, as the maximum threshold value.
As shown in FIG. 8A, the value of the parameter increases as the position of the slider 34S moves from left to right. The brightness of each pixel of slice 66 varies, and the result of edge detection depends on the value of the parameter. When the slider 34S is located at a relatively left position and the parameter value (threshold value) is small, as shown in FIG. 8B, the entire slice and the range of dust that is not a slice (not shown) are edge-extracted. When the slider 34S is in the center, most edges of the slice are extracted, as shown in FIG. 8C. As shown in FIG. 8D, when the slider 34S is on the right side, the edge of only the slice having a large difference in brightness from the background is extracted because the parameter value (threshold value) is large. According to these display results, the user sets the slider 34S at a desired position, and the detection unit 44 extracts an edge based on the value of the parameter determined by the position of the slider 34S.
The parameter is an example of the first parameter of the technique of the present disclosure.

In step 82B, the detection unit 44 calculates the direction of the luminance gradient in the image 34D of the wafer chip 64 shown on the left side of FIG. 9A using, for example, a sawbell filter, and as shown on the right side of FIG. 9A, each Luminance gradient direction table 34DE by storing the pixel identification data p1, p2, ... pn and the directions θ1, θ2, ... θn of the luminance gradient of the pixel identified by each identification data in correspondence with each other. To form.

The technique of the present disclosure is not limited to the method of calculating the direction of the luminance gradient by using the Sobel filter, and the direction of the luminance gradient may be calculated by using the Prewitt operator or the Roberts cross.

In step 82C, the detection unit 44 creates the filter 98. Specifically, as shown on the left side of FIG. 9B, first, the two parallel sides of the slice template 92 are extended vertically, respectively, and the two mask portions 94 and 96 are arranged above and below the slice template 92. .. The area of each of the upper and lower mask portions 94 and 96 is half the area of the slice template 92. The slice template 92 is set with V0 indicating that the brightness is lower than that of the mask portions 94 and 96, and the mask portions 94 and 96 are set with V1 indicating that the brightness is higher than that of the slice template 92. As described above, as shown on the right side of FIG. 9B, the filter 98 is determined by the luminance pattern (V1, V0, V1) of the mask portion 94, the slice template 92, and the mask portion 96. The slice template 92 defines a movement reference position 95 as a reference when moving the filter 98 in template matching described later.
The filter 98 is an example of the template of the technique of the present disclosure, the slice template 92 is an example of a region showing a sample section of the technique of the present disclosure, and the mask portions 94 and 96 are the background of the technique of the disclosure. It is an example of the region showing. The luminance (V1) with the mask portions 94 and 96 is an example of the value corresponding to the region showing the background of the technique of the present disclosure, and the luminance (V0) of the slice template 92 is the sample section of the technique of the present disclosure. It is an example of the value corresponding to the area shown.

Here, the mask portions 94 and 96 are arranged above and below the slice template 92, and the luminance patterns of the mask portion 94, the slice template 92 and the mask portion 96 are set to (V1 (large), V0 (small), V1 (large). )) Is used to correspond to the brightness of the actual slice 66 and the images above and below the slice. More specifically, as shown in FIG. 2B, since there is a background portion having a brightness higher than the brightness of the slice 93 above and below the slice 93, the brightness pattern of the mask portion 94, the slice template 92, and the mask portion 96. Is (V1 (large), V0 (small), V1 (large)), and the luminance pattern of the filter 98 corresponds to the luminance pattern of the slice 93 and the upper and lower portions of the slice 93.

As described above, the slice template 92 is formed from slices 93 having a shape relatively equal to a trapezoid.

From the above, when the image 34D of the wafer chip 64 is matched with the filter 98, the portion of the slice 93 can be detected from the image 34D of the wafer chip 64.

As shown in FIG. 2B, since the left and right portions of the slice 93 are the background in the image 34D of the wafer chip 64, it is conceivable that the filter 98 also arranges the mask portions on the left and right of the slice template 92. However, as shown in FIG. 2B, a plurality of slices 66 are often arranged side by side, and other slices are present on the left and right sides of the slice instead of the background. Therefore, the brightness on the left and right of the slice is smaller than the brightness of the background. Therefore, in the first embodiment, the mask portions are arranged above and below the slice template 92 to create the filter 98.

In step 82D, the detection unit 44 evaluates each pixel. Specifically, in the image 34D of the wafer chip 64, a matching process (template matching) is executed in the edge region extracted in step 82A by using the filter 98. In the matching process, for each pixel in the edge region of the image 34D of the wafer chip 64, the filter 98 arranged so that the movement reference position 95 is located in each pixel, and the portion in the image 34D where the filter 98 is arranged. The degree of similarity (a value representing the certainty that each pixel is a pixel of slice 66) is calculated (evaluated). Specifically, as shown on the left side of FIG. 9C, the filter 98 is moved along a predetermined direction 34DA with respect to the movement reference position 95. If the movement reference position 95 is not located in the edge region extracted in step 82A, the movement is continued without executing the matching process. When the filter 98 is moved and the movement reference position 95 is located in the area of the edge extracted in step 82A, the direction of the filter 98 is set to the direction θ of the brightness gradient of the pixel corresponding to the movement reference position 95. Then, the filter 98 is rotated to calculate the similarity. The above is performed for the entire image 34D of the wafer chip 64.

Specifically, the calculation of the similarity for each pixel is performed as follows. When the movement reference position 95 is located in the region of the edge extracted in step 82A, the evaluation region 102 in which the filter 98 is located is cut out from the image 34D of the wafer chip 64 as shown on the left side of FIG. Next, the pixel value of the cut out evaluation area 102 is binarized at the threshold value where 50% of the pixels of the cut out evaluation area 102 are white and the remaining 50% of the pixels are black. This is to correspond to the fact that 50% of the filter 98 (slice template 92) is white and the remaining 50% of the filter 98 (mask portions 94, 96) is black. Then, the degree of similarity between the luminance pattern of the binarized evaluation region 102A and the luminance pattern of the filter 98 is calculated.

In the evaluation area 102 shown on the left side of FIG. 10, since the luminance pattern of the binarized evaluation area 102A (right side of FIG. 10) and the luminance pattern of the filter 98 are almost the same, the similarity value (score) is used. Is calculated as a relatively large value, 0.982 (on the right side of FIG. 10).

From the scores calculated for each pixel, as shown on the right side of FIG. 9C, the identification data p1, p2, ... pn of each pixel and the scores s1, s2, ... Sn of the pixels identified by each identification data. The score table 34DM is formed by correspondingly storing the above.

As described above, by using the filter 98 based on the light-dark distribution of the slice part and the background part and performing the calculation using the binarized value of the brightness value, the robustness of detection against the fluctuation of the brightness is achieved. Is getting. For example, even if the slice is difficult to detect, such as a slice 66 having a relatively high brightness or a slice 66 having a variation in brightness inside the slice, if the condition that the brightness of the slice portion is smaller than that of the background portion is satisfied. , Can be detected.
The matching process (template matching) is performed using the filter 98 in order to detect the slice, but the process performed to detect the slice is limited to the matching process (template matching) using the filter 98. Instead, it may be found, for example, using machine learning (eg, Deep Learning).

In step 82E of FIG. 7, the detection unit 44 determines the position of each slice 66 in the image 34D of the wafer chip 64. Specifically, the detection unit 44 positions the origin (movement reference position 95) of the slice template 92 on the maximum pixel of the score extracted from the score table 34DM (see the right side of FIG. 9C), and sets the brightness gradient direction table 34DE. From, slice candidates 112 are placed as shown in FIG. 11A and slices 114 are set as shown in FIG. 11B along a direction determined by the direction of the intensity gradient corresponding to the highest pixel of the score.

Next, for all pixels other than the largest pixel whose value (score) indicating the certainty that each pixel of the image 34D of the wafer chip 64 shown in FIG. 12A is a pixel of the slice 66 is equal to or higher than the score threshold value, the score is set. In order from the largest pixel, it is checked whether it is a pixel of the slice, and the slice is confirmed according to the result of the check. Specifically, as shown in FIG. 12B, the slice candidate 116 is arranged in the pixel having the second highest score among the plurality of pixels having the score threshold value or more by the method described with reference to FIGS. 11A and 11B. When the arranged slice candidate 116 overlaps with the already detected slice 114 by a predetermined value or more, the slice candidate 116 is not detected as a slice as shown in FIG. 12C. Next, as shown in FIG. 12D, the slice candidate 118 is arranged in the pixel having the third highest score among the plurality of pixels having a score threshold value or higher by the method described with reference to FIG. When the arranged slice candidate 118 overlaps with the detected slice 114 by less than a predetermined value, the slice candidate 118 is determined as the slice 120 as shown in FIG. 12E. The above is performed for pixels above the score threshold.

When the slice confirmation process of step 82E of FIG. 7 is completed, the slice detection process proceeds to step 84 of FIG.
Instead of the automatic detection in step 82 of FIG. 3, a process of detecting all slices by the user may be executed.

In step 84 of FIG. 3, the adjusting unit 46 manually adjusts the determined slice.

Specifically, in step 84A, the erroneous detection slice is deleted. Specifically, it was detected as a slice by the slice detection process, but when the user sees the displayed detection result and determines that it is not a slice, the slice is deleted as a false detection.

Further, in step 84B, the user fine-tunes the position of the detected slice.

In step 86, the ordering unit 48 performs slice ordering. That is, the user specifies in what order the slices are to be stacked in order to construct the three-dimensional data of the sample via the input unit 36. In step 86, the ordering unit 48 detects the input of the order by the user. For example, when the user specifies the order of the slice columns 132 and 134 and the order of each slice of the slice columns 132 and 134 via the input unit 36 as shown in FIG. 13, the ordering unit 48 uses the slice sequence. The order of 132 and 134 and the order of each slice of the slice rows 132 and 134 are detected.

In step 88, the designation unit 50 designates a slice to be excluded from the three-dimensional structure data (3D data). If the user determines that the image 34D of the wafer chip 64 is not a false positive but a slice, but is not suitable for being deformed and used as data (3D data) of the three-dimensional structure of the sample, the user decides. Click the slice to be deleted via the input unit 36. In step 88, the designation unit 50 detects the input by the user and deletes the data of the detected slice.

As shown in FIG. 14A, the X-axis and the Y-axis are defined on the wafer 60 with reference to the three points 62A, 62B, and 62C to which the absolute position marker is attached.
In step 90, the determination unit 52 determines the position of each corner of each slice 66 with respect to the three points 62A, 62B, 62C to which the absolute position marker is attached, that is, the coordinates on the X-axis and the Y-axis. do. Specifically, the determination unit 52 outputs the coordinate information of (x1, y1) for the point P1, (x2, y2) for the point P2, (x3, y3) for P3, and (x4, y4) for P4. For example, (292,80), (310,114), (378,77), (335,57) are determined.

In step 92, the transmission unit 54 transmits the position data of the slice 66 with respect to the three points 62A, 62B, 62C to which the absolute position marker is attached to the electron microscope 14 via the communication interface 38.

The position data of the slice 66 is not limited to the coordinate information of the points at each corner of the slice, and first, the coordinates of the two representative points, for example, (292,80) and (310,114) may be used. In 2, the coordinates of one representative point and the angle with respect to the line connecting the two points 62B and 62C to which the absolute position marker is attached may be, for example, (292,80), 61 degrees.

Further, the position data of the slice 66 is not limited to the reference of the three points 62A, 62B, and 62C to which the absolute position marker is attached, and the absolute position marker is added to the four points of each corner of the wafer chip 64. The determination is made based on the three points 62A, 62B, and 62C, and the position data of the slice 66 may be based on the points of the three corners of the wafer chip 64. If the four points of each corner of the wafer chip 64 are determined based on the three points 62A, 62B, and 62C to which the absolute position marker is attached, the slice 66 obtained based on the points of the three corners of the wafer chip 64. This is because the position data of can be converted into a value based on the three points 62A, 62B, and 62C to which the absolute position marker is attached.

As described above, the image processing apparatus 16 of the first embodiment can acquire the coordinate information of each slice.

As described above, the points to which the three points 62A, 62B, and 62C are given as the absolute position markers are common reference positions in the optical microscope 12, the electron microscope 14, and the image processing apparatus 16. Therefore, when the position of the slice 66 is specified with reference to the three points 62A, 62B, 62C to which the absolute position marker is given, the position of the specified slice 66 is determined by the optical microscope 12, the electron microscope 14, and the image processing device 16. It can be used in common.

Therefore, in step 90, when the image processing apparatus 16 transmits the data of the position of the slice 66 specified with reference to the three points 62A, 62B, 62C to which the absolute position marker is attached to the electron microscope 14, the electron microscope 14 receives the data. The position of each slice 66 can be specified with reference to the three points 62A, 62B, and 62C to which the absolute position marker is attached.

By the way, the portion of the surface of the slice to be observed (ROI (Region Of Interest)) is relatively the same position for each slice, but in a conventional electron microscope, the user slices each slice in the image of the wafer chip. The points to be observed on the surface of the surface were individually specified. When the electron microscope acquires the data of the part to be observed in one slice, the wafer is moved so that the individually specified part of another slice is located in the observation range, and the data of the specified part of each slice is obtained. I had acquired it.

As described above, in the conventional technique, since the user individually specifies the points to be observed that are relatively at the same position for each slice for each slice, the labor of the user is excessive.

In the first embodiment, the position of each slice 66 is specified so that it can be commonly used in the optical microscope 12, the electron microscope 14, and the image processing apparatus 16, and each slice 66 has a size and a shape. Is the same, so specify the part you want to observe in one slice 66 with the electron microscope 14, for example, if the part you want to observe is at one point, click it, and if it is an area, enclose it with a rectangle or the like using a mouse. When specified by, the portion to be observed of all the other slices 66 can be automatically specified based on the position of the designated portion. Therefore, in the electron microscope 14, when the data of the designated portion to be observed in one slice 66 is acquired, another slice is obtained from the specified position of each slice 66 and the position of the portion to be observed in one slice 66. The wafer 60 can be moved so that the corresponding portion to be observed is located in the observation range, and the data of the portion to be observed of each slice can be automatically acquired.

By designating the part to be observed in one slice 66 in this way, the data of the part to be observed in all the other slices 66 can be automatically acquired, so that the user's labor can be saved in the conventional electron microscope apparatus. It can be less labor than required.

From the data of the part to be observed obtained for each slice by the electron microscope 14, the data of the three-dimensional structure of the part to be observed is constructed according to the order of the slices determined by the user or automatically.

Then, the three-dimensional data of the sample acquired by the optical microscope 12 and the data of the three-dimensional structure of the sample constructed by the electron microscope 14 may be combined and displayed. For example, the entire sample may be displayed, and only a desired portion of the sample may be displayed in different colors.

(Second embodiment)
Next, a second embodiment of the technique of the present disclosure will be described. Since the configuration of the second embodiment is the same as the configuration of the first embodiment, the description thereof will be omitted.

Next, the operation of the second embodiment will be described. The operation of the second embodiment is different in that the semi-manual detection process (FIG. 15) described later is executed after the slice determination process of step 82E of FIG. 7. The detection unit 44 (see FIG. 1B) executes the semi-manual detection process.

In the first embodiment, the slice is automatically detected by the auto detection process, but there may be a case where the slice is not detected. In the second embodiment, the leaked slice is detected with the assistance of the user.

In step 152 of FIG. 15, the detection unit 44 detects a clicked portion of the slice to be detected that has been missed from the detection. Specifically, in the image 34D shown in FIG. 16A in which the slice detected by the auto detection process is displayed, the user clicks the point 152P1 at the center of the slice recognized as a slice via the input unit 36. The detection unit 44 detects the clicked point 152P1.
The image shown in FIG. 16A is an example of a second image of the technique of the present disclosure.

In step 154, the detection unit 44 cuts out a region in the vicinity of the clicked point. Specifically, the detection unit 44 cuts out a square area around the clicked point and sets it as a search area 154R. One side of the square area is determined, for example, from the diagonal length of the slice template 92 so that the slice template 92 fits within the area. Specifically, it is in the range of 1.0 times to 2.0 times. The shorter it is, the better the accuracy. The longer it is, the less accurate the click is required of the user. For example, a square region that is 4/3 times the diagonal of the slice template 92 may be used as the search region. As shown in FIG. 16A, when the point 152P1 of the slice recognized as the leaking slice is clicked, as shown in FIG. 16B, the square region 154R centered on the clicked point 152P1 is cut out as a search region. ..

In step 156, the detection unit 44 sets a parameter for edge extraction. Specifically, as shown in FIG. 17A, a parameter for edge extraction is set from the slice brightness of a predetermined region including the point 152P1 clicked in step 152 and the background brightness. The parameter for edge extraction may be, for example, a threshold in the Canny method. Specifically, as shown in FIG. 17B, the average value of the brightness of the pixels in the predetermined area 152PR including the clicked point 152P1 is calculated as the slice brightness. Further, as shown in FIG. 17C, a histogram of the brightness of the entire image of the wafer chip is formed, and the mode brightness value in the histogram is selected as the background brightness. In the example shown in FIG. 17C, the total number of pixels (Count) is 79500, and the luminance value of the mode (10679) of luminance is 151. For example, when the edge is extracted by using the Canny method, a value corresponding to the value of the difference between the slice brightness and the background brightness is set as a parameter (threshold value) for edge extraction. The value corresponding to the value of the difference between the slice brightness and the background brightness may be the value itself of the difference between the slice brightness and the background brightness, and the value of 1/4 or 1/4 of the difference between the slice brightness and the background brightness. A value of 2 may be used, and is appropriately determined according to the resolution and magnification of the image of the wafer chip. The parameter for edge extraction in the second embodiment is set to a value more suitable for the search region 154R than the parameter for edge extraction in the first embodiment to facilitate edge extraction.
The parameter for edge extraction in step 156 is an example of the second parameter of the technique of the present disclosure.

In step 158, the detection unit 44 extracts edges by using, for example, the Canny method, as in the auto-detection process. As described above, in step 156, a value corresponding to the difference between the slice brightness and the background brightness is set as a parameter (threshold value) for edge extraction. Therefore, as shown in FIG. 17D, the slice that was not detected by the auto detection. Edge 158E0 is extracted.

In step 156, the detection unit 44 sets the edge extraction parameter to a value corresponding to the difference between the slice brightness and the background brightness, but the technique of the present disclosure is not limited to this, and a predetermined value may be used. good.

In step 160 of FIG. 15, the detection unit 44 calculates (evaluates) a similarity value (score) indicating the certainty that each pixel is a pixel of slice 66 for each pixel of the search area 154R cut out in step 154. do. Specifically, in the search area 154R shown in FIG. 18A, the slice template 92 is moved, and for each pixel, the slice template 92 shown in FIG. 18B is arranged so that the movement reference position 95 corresponds to each pixel. When the slice template 92 is moved and the movement reference position 95 is located in the area of the extracted edge, the orientation of the slice template 92 is set to the direction θ of the brightness gradient of the pixel corresponding to the movement reference position 95. , The slice template 92 is rotated to calculate the similarity. From the overlap between the slice template 92 and the extracted edges, a similarity value (score) representing the certainty that each pixel is a pixel of the slice 66 is calculated, and a score table is created.
The slice template 92 used in step 160 of FIG. 15 is an example of a template determined based on the shape of the sample section of the technique of the present disclosure.

In step 162, the detection unit 44 determines the slice. Specifically, using the score table in step 160, the slice template 92 is arranged so that the movement reference position 95 corresponds to the pixel having the maximum score, and the slice is confirmed. In this case, since one leaked slice is selected by the user, one candidate is always detected. The method of determining the overlap is the same as the determination of slices in the auto detection process (step 82E in FIG. 7). The detection result information (confirmed slice) used for the overlap determination is common to that of the automatic detection.

As shown in FIG. 19A, the slice 166N is determined for the square search area 154R centered on the clicked point, as shown in FIG. 19B.

As described above, in the second embodiment, the user assists by having the user click the leaked slice for the slice leaked from the detection by the auto detection process. Leaked slices are detected by setting the parameters so that edges can be easily extracted rather than the parameters for edge extraction in the auto detection process. Therefore, in the second embodiment, the slice data required as the data of the three-dimensional structure of the sample can be detected without omission.

Further, in the second embodiment, the similarity value (score) for each pixel of the search area 154R, which is known to always have one leaked slice, is determined by the shape of the slice template 92 and the auto detection process. It is calculated from the degree of overlap with the extracted edges by making the parameters easier to extract than the parameters for edge extraction in. Then, the slice template 92 is arranged so that the movement reference position 95 corresponds to the pixel having the maximum score, and the slice is confirmed. Therefore, in the second embodiment, the leaked slice can be appropriately detected.
In the second embodiment, the luminance pattern of the filter 98 is set in the same manner as in the first embodiment in order to calculate the value representing the certainty that each pixel of the search region 154R is a pixel of the slice 66. You may use it.

Next, a modified example of the technique of the present disclosure will be described. Since the configuration of each of the following modifications is the same as the configuration of the first embodiment, the description thereof will be omitted. Further, since the operation of each modification has the same part as that of the first embodiment, only the different part will be described.

(First modification)
In the first embodiment and the second embodiment, the user selects a slice 93 having a relatively correct shape as a slice template from a plurality of slices 66 in the image of the wafer chip 64 shown in FIG. 2B. Then, the corners of the slice 93 on the image are clicked in order to specify 4 points (or 5 points). The technique of the present disclosure is not limited to this, and 4 points (or 5 points) of slice templates may be specified as follows.

First, the user inputs the coordinate values of the four (or five) slice templates via the input unit 36.

Second, when the user specifies to trace the contour of one slice from the image via the input unit 36, the critical shape is detected, and the range including one slice is specified. , The edge is detected by using the brightness value, etc., and the shape is detected. Then, the coordinates of 4 points (or 5 points) are specified from the detected shape.

(Second modification)
In the first embodiment and the second embodiment, the image of the wafer chip 64 is subjected to matching processing using the filter 98, so that each pixel of the image of the wafer chip 64 is sliced 66. Calculate (evaluate) a value (similarity) that represents the certainty of being a pixel.

However, the slices are not scattered on the wafer chip 64, but are arranged in a row. Therefore, it is not necessary to perform matching processing on the entire image of the wafer chip 64.

Therefore, as shown in FIG. 20, the second modification is performed in step 82G in order to limit the area to be matched before step 82A of the auto detection process (FIG. 7) of the first embodiment. The detection unit 44 executes the process of designating the search area. The search area designation process in step 82G is the search position-limited user assist process in step 81 of FIG. The process of designating the search area in step 82G may be performed after the process of edge extraction in step 82A.

In step 82G, specifically, in the image 34D of the wafer chip 64 shown in FIG. 21A, as shown in FIG. 21B, the movement reference position 95 is set for the matching process designated by the user via the input unit 36. The region 82G1 to which the filter 98 is moved to the reference is detected. Then, as in the first embodiment, steps 82C to 82E are executed. In the matching process in step 82D, the similarity is calculated while moving the filter 98 only in the region of the edge extracted in step 82A, which is also the region 82G1, and the slice is determined as shown in FIG. 21C.

The process of designating the search area is not limited to the example shown in FIG. 21B, and as shown in FIG. 21D, the user sets the area 82GR1 including the slice via the input unit 36. In step 82G, the detection unit 44 detects the region 82GR1 including the slice specified by the user via the input unit 36. Then, as in the first embodiment, steps 82C to 82E are executed. The matching process in step 82D is executed only in the region of the edge extracted in step 82A and in the designated region 82GR1. Therefore, the edge region outside the designated region 82GR1 is ignored.

Further, as shown in FIG. 21E, the user designates end points 82GP1, 82GP2, and 82GP3 of continuous slices via the input unit 36. In step 82G, the detection unit 44 detects the end points 82GP1, 82GP2, 82GP3 designated by the user via the input unit 36, and interpolates between the end points with a straight line. Then, as in the first embodiment, steps 82C to 82E are executed. The matching process in step 82D is executed only in the region of the edge extracted in step 82A and the region determined by the endpoints 82GP1, 82GP2, 82GP3 (for example, the region interpolated by a straight line between 82GP1, 82GP2, 82GP3). The distance between the endpoint 82GP1 and the endpoint 82GP2 is divided by the width of the slice template 92 to obtain the number of slices between the endpoint 82GP1 and the endpoint 82GP2. In the matching process, the edge region extracted in step 82A is obtained. In addition, in the area between the end points 82GP1 and the end point 82GP2, the matching process may be executed only in the area allocated according to the number of slices. In the example shown in FIG. 21E, the user specifies the end points 82GP1, 82GP2, and 82GP3 via the input unit 36, but the technique of the present disclosure is not limited to this. First, the range of the extracted edges may be approximated to a straight line to automatically detect the end points 82GP1 to the end points 82GP3. Secondly, the number of slices may be obtained by extracting the triangular portion at the tip of the trapezoid within the range of the extracted edges and counting the number of the extracted triangles.

In the second modification, by limiting the search range of the slice in a region or a direction, the quality accuracy can be improved and the processing time can be shortened. The search range of the slice is not limited to the edge, the edge, and the user-specified range.
The range including the region 82G1, the region 82GR1, the region determined by the end points 82GP1, 82GP2, and 82GP3 is an example of the partially specified region of the technique of the present disclosure.

(Third modification example)
In the setting of the edge extraction parameter of the first embodiment (step 82A in FIG. 7), the user first sets the slider 34S at a desired position. In step 82A, the position of the slider 34S is detected, and the edge extraction parameter corresponding to the detected position is set. The technique of the present disclosure is not limited to the user setting the slider 34S at a desired position in this way to set the parameters of edge extraction. For example, a plurality of parameters are determined in advance. For example, as shown in FIG. 22, 250, 200, 150, 100 are determined. Edges are extracted with each of the four parameters, and the extracted edges are displayed in a list. The first display area 34D1 shows the edge extraction result when the parameter is 250. In the second display area 34D2, the edge extraction result in the case of the parameter 200 is shown. In the third display area 34D3, the edge extraction result in the case of the parameter 150 is shown. In the fourth display area 34D4, the edge extraction result in the case of the parameter 100 is shown. The user clicks on a display area in which the four display results are preferred. For example, click the third display area 34D3. In step 82A, 150 is set as a parameter.

Further, the present invention is not limited to determining a plurality of parameters in advance, and the parameters are automatically and gradually changed, edges are extracted each time the parameters are changed, and the extracted results are displayed. For example, the parameters are gradually changed every 20 from 250 to 230, 210, 190, .... Edge extraction is performed with each changed parameter, and the extracted result is displayed. The user clicks the screen when the extraction result determined to be appropriate is displayed. In step 82A, the parameter when clicked is set.

In the third modification, since the edge extraction process has already been executed, the edge extraction process in step 82A is omitted.

(Fourth modification)
The process of extracting edges is not limited to the Canny method.

(Fifth variant)
In the first embodiment, as a template for template matching, a filter 98 determined by a luminance pattern (V1, V0, V1) between the mask unit 94, the slice template 92, and the mask unit 96 is used. Further, in the second embodiment, the slice template 92 is used as a template for template matching. The techniques of the present disclosure are not limited to this.

First, in the template matching of the auto detection of the first embodiment, the slice template 92 is used as a template, the template matching of the semi-manual detection of the second embodiment is executed, and the template matching of the second embodiment is performed. In the template matching of semi-manual detection, the filter 98 may be used as a template to perform the template matching of the auto detection of the first embodiment.

Secondly, in the template matching of the first embodiment and the second embodiment, the filter 98 may be used as a template to perform the template matching of the first embodiment.

Thirdly, in the template matching of the first embodiment and the second embodiment, the slice template 92 may be used as a template to perform the template matching of the second embodiment.

Fourth, in each of the template matching of the first embodiment and the second embodiment, template matching using the filter 98 and template matching using the slice template 92 are performed. The average of the scores of each pixel obtained by each template matching is calculated for each pixel. The slice may be specified from the average of the scores of each pixel obtained for each pixel. It should be noted that the above average is not a simple average, and the average of the scores in the auto detection has a greater influence on the result of the template matching score using the filter 98, and the average of the scores in the semi-manual detection is the slice template 92. The score of template matching using can be made to have a greater influence on the result.

(Additional note)
From the above, the following programs are proposed.
(Appendix 1)
Computer,
In the first image obtained by photographing a substrate on which a plurality of sample sections are arranged, the detected sample section is used by using a detection unit for detecting at least one sample section and a predetermined reference position. A program that acts as a determinant to determine the position of.
(Appendix 2)
The program according to Appendix 1, wherein the sample section is obtained by slicing a sample.
(Appendix 3)
The program according to Appendix 1 or 2, wherein the detection unit detects the sample section by performing template matching using the template of the sample section.
(Appendix 4)
The detection unit
For each pixel of the first image, the direction of the luminance gradient is detected.
The template matching is performed for each of the pixels based on the direction of the brightness gradient.
The program described in Appendix 3.
(Appendix 5)
The detection unit
By performing the template matching, for each of the pixels, a score indicating the certainty that each pixel is a pixel of the sample section is detected.
Detecting the sample section based on the score,
The program described in Appendix 3 or 4.
(Appendix 6)
The detection unit identifies the edge of the sample section using the first parameter in the first image.
The program according to any one of Supplementary note 1 to Supplementary note 5, wherein the sample section is detected using the specified edge.
(Appendix 7)
The program according to Appendix 6, wherein the detection unit detects the sample section using the region specified in the first image and the specified edge.
(Appendix 8)
The program according to any one of Supplementary note 1 to Supplementary note 7, wherein the detection unit further detects a sample by using the position selected in the second image showing the detection result of the sample section.
(Appendix 9)
The program according to Appendix 8, wherein the detection unit detects the sample section in a part of the region of the second image determined based on the selected position.
(Appendix 10)
The detection unit identifies the edge of the sample section using the first parameter in the first image, and identifies the edge using the second parameter in a part of the region of the second image. ,
The program described in Appendix 9.
(Appendix 11)
The second parameter is a value that makes it easier to identify the edge than the first parameter.
The program described in Appendix 10.
(Appendix 12)
The program according to any one of Supplementary note 8 to Supplementary note 11, wherein the detection unit further detects a sample section in the second image by performing template matching using the template of the sample section.
(Appendix 13)
The template is
A region showing the sample section, a region showing the background, and a region showing the background.
The value corresponding to the region showing the sample section and the value corresponding to the region showing the background,
The program according to any one of Supplementary note 3 to Supplementary note 5 and Supplementary note 12, wherein the program has.
(Appendix 14)
The template is the program according to any one of Supplementary note 3 to Supplementary note 5 and Supplementary note 12, which is determined based on the shape of the sample section.
(Appendix 15)
Computer,
Using the first parameter in the first image, a predetermined shape is detected and
A detection unit that detects the predetermined shape using the second parameter in a part of the region of the second image including the position selected in the second image showing the detected result, and a predetermined A program that functions as a determination unit for determining the position of the detected predetermined shape by using the reference position of the above.
(Appendix 16)
Computer,
A program that functions as a detection unit that detects a predetermined shape in the first image and a display unit that displays a second image showing the detected result by using the first parameter.
The detection unit
A program that detects the predetermined shape using the second parameter in a part of the area of the second image including the position selected in the second image.

10 CLE M system 12 Optical microscope 14 Electron microscope 16 Image processing device 20 Computer 42 Creation unit 44 Detection unit 46 Adjustment unit 48 Ordering unit 50 Designation unit 52 Determination unit 54 Transmission unit 60 Wafer 64 Wafer chip 92 Slice template 94, 96 Mask unit 98 filter

Claims (18)

In the first image obtained by photographing a substrate on which a plurality of sample sections are arranged, a detection unit for detecting at least two of the sample sections and a detection unit.
A determination unit that obtains information on the position of each detected sample section using a predetermined reference position, and
A data processing unit that builds three-dimensional data of a sample by stacking images of each sample section acquired by an electron microscope based on the position information.
A device equipped with. Within one sample section of each of the sample sections, an area designation unit that accepts the user's designation of the observation area used for constructing the three-dimensional data is provided.
The determination unit is an observation region located at a relatively same position in each sample section based on the information on the position of the observation region received by the region designation unit and the information on the position of each sample section. Identify the location and
The data processing unit constructs the three-dimensional data by stacking images of the observation region in each sample section acquired by the electron microscope.
The device according to claim 1. It is provided with an ordering unit that receives instructions from the user regarding the order in which each sample section is laminated.
The data processing unit constructs the three-dimensional data by stacking images of the sample sections in an order based on the instruction received by the ordering unit.
The device according to claim 1 or 2. The apparatus according to any one of claims 1 to 3, wherein the sample section is obtained by slicing a sample. The apparatus according to any one of claims 1 to 4, wherein the detection unit detects the sample section by performing template matching using the template of the sample section. The detection unit
For each pixel of the first image, the direction of the luminance gradient is detected.
The template matching is performed for each of the pixels based on the direction of the brightness gradient.
The device according to claim 5. The detection unit
By performing the template matching, for each pixel, a score indicating the certainty that each pixel is a pixel of the sample section is detected.
Detecting the sample section based on the score,
The device according to claim 5 or 6. The detection unit identifies the edge of the sample section using the first parameter in the first image.
The apparatus according to any one of claims 1 to 7, wherein the sample section is detected using the specified edge. The apparatus according to claim 8, wherein the detection unit detects the sample section using the region specified in the first image and the specified edge. The apparatus according to any one of claims 1 to 9, wherein the detection unit further detects a sample by using a position selected in a second image showing a detection result of the sample section. The apparatus according to claim 10, wherein the detection unit detects the sample section in a part of the region of the second image determined based on the selected position. The detection unit identifies the edge of the sample section using the first parameter in the first image, and identifies the edge using the second parameter in a part of the region of the second image. ,
The device according to claim 11. The second parameter is a value that makes it easier to identify the edge than the first parameter.
The device according to claim 12. The apparatus according to any one of claims 10 to 13, wherein the detection unit further detects a sample section in the second image by performing template matching using the template of the sample section. The template is
A region showing the sample section, a region showing the background, and a region showing the background.
The value corresponding to the region showing the sample section and the value corresponding to the region showing the background,
The apparatus according to any one of claims 5 to 7, and 14. The apparatus according to any one of claims 5 to 7, and 14, wherein the template is determined based on the shape of the sample section. The detection unit detects at least two of the sample sections in the first image obtained by photographing the substrate on which a plurality of sample sections are arranged.
The determination unit obtains information on the position of the detected sample section using a predetermined reference position.
The data processing unit builds three-dimensional data by stacking images of each sample section acquired by the electron microscope based on the position information.
Method. On the computer
A procedure for detecting at least two sample sections in a first image obtained by photographing a substrate on which a plurality of sample sections are arranged, and
A procedure for obtaining information on the position of the detected sample section using a predetermined reference position, and a procedure for obtaining information on the position of the detected sample section.
A procedure for constructing three-dimensional data by superimposing images of each sample section acquired by an electron microscope based on the position information, and
A program to execute .

Images