JP2008116206A - Apparatus, method, and program for pattern size measurement - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a constitution capable of improving accuracy in measuring the size of a prescribed section of a given pattern. <P>SOLUTION: The constitution comprises an approximate line acquisition means for determining approximate lines by prescribed operations on the basis of point groups constituting the given pattern; a reference point acquisition means for determining reference points by prescribed operations on the basis of the approximate lines; a length measuring position acquisition means for determining length-measuring positions by prescribed operations on the basis of the reference points; and a size measuring means for measuring the size of a prescribed section of the pattern at the length-measuring positions. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a pattern dimension measuring apparatus, method, and program, and more particularly to a pattern dimension measuring apparatus, method, and program for measuring a dimension of a predetermined portion of a given pattern.

  A scanning electron microscope for length measurement, a so-called CD-SEM, has a function of imaging a predetermined location in a semiconductor wafer and measuring the length of the same location, and a manufacturing process of a recording / reproducing head of a semiconductor device or a hard disk device Etc. are widely used.

  The general method of this device is as follows.

  First, coordinates to be imaged on a semiconductor wafer, a so-called teacher image serving as a reference, a measurement position in an image related to the image of the semiconductor wafer, and an edge detection algorithm are registered in advance as measurement recipes.

  Next, the imaging position is moved to a predetermined coordinate on the semiconductor wafer by roughly moving the stage on which the semiconductor wafer is placed. Then, by comparing the scanning information obtained by imaging with the previously registered teacher image by template matching, the image range related to the length measurement is determined.

  Finally, an edge included in the image is detected using scanning information at a predetermined position on the screen on which the image range thus determined is projected. Then, the distance between the detected edges is measured, that is, the length is measured. In the length measurement, scanning information before compression processing for projecting an image on the screen is used, and therefore the length measurement accuracy is generally high.

  Further, Patent Document 1 discloses a technique for measuring a length with high accuracy when an image obtained by imaging with a scanning electron microscope or the like is used as a target.

  In the technique disclosed in Patent Document 1, in a microscope image obtained by imaging through a scanning microscope or the like, the pattern shape is specified by approximating all the sides of the pattern included in the image with straight lines. And the specific location of the said pattern is measured by calculating the distance between two points, such as a specific point calculated from the intersection between the said straight lines.

  In an image of a scanning electron microscope or the like, generally, the pattern included in the image tends to be imaged with a dull angle. However, according to the method of the above-mentioned Patent Document 1, the task of specifying the measurement point by the subjectivity of the measurer is eliminated. Therefore, it is possible to reduce the length variation.

  However, in the CD-SEM length measurement, the positioning position positioning accuracy depends on the template matching accuracy. Here, in general, when a pattern with few features is a length measurement target, the accuracy of template matching is likely to be lowered, and the positioning accuracy of the length measurement position may be lowered.

In addition, in the case of measuring the length of a part that has no features in the pattern included in the microscopic image, for example, the part that is scheduled to be processed in a later manufacturing process, the method for specifying the pattern shape by the disclosed technique of Patent Document 1, It is considered that it is not always easy to accurately determine the measurement position.
JP 2002-350127 A

  The present invention has been made in view of such a situation, and when measuring a specific portion in a pattern included in a microscope image obtained by the CD-SEM, the electron microscope, or the like, the specific portion It is an object of the present invention to provide a configuration capable of accurately determining the length measurement accuracy and improving the length measurement accuracy of the same portion.

  In the present invention, an approximate line is obtained from a point group constituting a given pattern, a reference point is obtained from the approximate line, and a length measurement position is obtained from the reference point.

  As a result, the length measurement position can be obtained with high accuracy.

  According to the present invention, the measurement position can be determined with high accuracy by a relatively simple method. Therefore, when the manufactured product is inspected using the image obtained by the CD-SEM or the like, high accuracy and high speed can be obtained. Inspection work can be performed.

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

  An embodiment of the present invention is an image length measuring device having a function of determining a measurement position with high accuracy with respect to a microscope image obtained by a scanning electron microscope, a CCD camera or the like.

  That is, the image length measuring device according to the embodiment of the present invention has a function of obtaining a point serving as a length measurement reference from a reference line obtained by approximating the sides of the pattern in the microscope image. As a result, the positioning accuracy of the length measurement position can be effectively improved.

  The image length measuring device according to the first embodiment of the present invention to be described later is a device for measuring a dimension of a pattern included in the image at a predetermined position in a microscope image, that is, a length measuring method. Detecting a point group that forms each side part of the constituent, obtaining an approximate line of each side part constituting the contour of the pattern from the point group, and determining a reference point from the same approximate line or one or more intersections of the same approximate line It has a function of obtaining a length measurement position from the reference point and measuring a dimension between edges, which is a contour of the same pattern extracted by edge detection at the length measurement position, that is, length measurement.

  In the image length measuring device according to the second embodiment of the present invention, which will be described later, when detecting a point group forming each side part constituting the outline of the pattern, or at each side part forming the pattern outline from the point group. When obtaining the approximate line, the area of the image for extracting the approximate line is limited in advance according to the type of the microscope image to be measured. Then, by performing template matching between a so-called teacher image prepared in advance and an image in the restricted area, a positional deviation between the microscope image and the teacher image is detected. Based on the detection result, in order to correct the detected positional deviation, a function of moving an image region for extracting the approximate line on the microscope image is provided.

  In the image length measuring apparatus according to the third embodiment of the present invention, it is prepared in advance when detecting a point group forming each side part constituting the contour of the pattern and when measuring a pattern dimension by the edge detection. Template matching is performed between the plurality of teacher images and the microscope image. Then, a length measurement position on the microscope image is determined based on a comparison table prepared in advance, and an edge detection algorithm (described later) at the time of the length measurement is selected. Thereafter, the dimension of the pattern is measured, that is, the length is measured using the optimum edge detection algorithm according to the selection for the measurement position according to the determination.

  In the image length measuring device according to the fourth embodiment of the present invention, when detecting a point group forming each side portion constituting the contour of the pattern, template matching is performed between a teacher image prepared in advance and the microscope image. And the correlation between the two images is calculated. Then, when the calculated correlation between the images is larger than a predetermined threshold, the subsequent procedure is performed.

  In the image length measuring device according to the fifth embodiment of the present invention, a means for electronically storing a microscope image taken by a microscope is provided, and the dimension of the pattern included in the image is measured at a predetermined position, that is, the length is measured. In this case, it is detected that a microscope image is obtained by imaging and stored in the storing means. In response to the detection, a predetermined length measurement operation is started with respect to the stored microscope image, and the result of the operation is output.

  According to the image length measuring apparatus of each of the above embodiments of the present invention, a microscope image obtained by a scanning electron microscope or the like is obtained from an approximate line of each side that forms the contour of a pattern included in the image. Since the measurement position is determined based on the determined point, the variation in the measurement position can be effectively reduced. As a result, it is possible to effectively reduce variations in length measurement results due to variations in length measurement positions.

  Further, in the pattern included in the microscope image, the position corresponding to the part to be processed in the subsequent manufacturing process in the corresponding manufactured product, and the feature in the pattern in the microscope image is particularly noticeable in shape Even in the case of measuring the length of a position where there is no position, the length measurement position can be determined with high accuracy. Therefore, it is possible to realize a more advanced length measurement operation, and efficiently inspect the manufactured product in the manufacturing process of the recording / reproducing head of the semiconductor device and the hard disk device, thereby improving the product yield. The manufacturing cost can be greatly reduced.

  The function of the image length measuring device according to each embodiment of the present invention will be described below in detail with reference to the drawings.

  First, the image length measuring apparatus according to the first embodiment of the present invention will be described.

  FIG. 1 is a diagram for explaining the function of the image length measuring device 50 according to the first embodiment of the present invention. Here, as an example, it is assumed that the image measuring device 50 is configured by a personal computer.

  As shown in FIG. 1, the image length measuring device 50 is connected to the CD-SEM 100 and receives a captured image of a product to be inspected, that is, a microscope image, from a personal computer 110 that controls the CD-SEM 100. Then, the image length measuring device 50 performs image processing on the transferred microscope image, thereby measuring the dimension of a predetermined portion of the inspection target product, that is, measuring the length. In this manner, whether or not the manufactured product is within a predetermined allowable range is determined based on the information on the dimensions of the predetermined part of the inspection target product.

  FIG. 2 is a functional block diagram for explaining the function of the image length measuring device 50, and FIG. 3 is an operation flowchart for explaining the function of the image length measuring device 50.

  As shown in FIG. 2, the image length measuring apparatus 50 extracts an image acquisition unit 511 that receives a captured image of a product to be inspected from the personal computer 110 of the CD-SEM 100, that is, a microscope image, and an edge of a pattern included in the microscope image. Edge extracting unit 512, approximate line acquiring unit 513 for acquiring an approximate line from the extracted edge, reference point acquiring unit 514 for determining a reference point from the acquired approximate line, and acquiring a measurement position from the determined reference point Length measurement position acquisition unit 515, an edge detection unit 516 that detects an edge at the measurement position acquired by the measurement position acquisition unit 515 in the microscope image, and measures a distance between the detected edges, that is, a measurement for measuring the length. A long part 517 and a result output part 518 for outputting a length measurement result are included.

  In the case of this example, each of these units 511 to 518 is realized by the CPU of a personal computer constituting the image measuring device 50 executing a predetermined program.

  Further, here, as an example of the microscopic image of the inspection target product, that is, the microscopic image of the length measurement target, a scanning electron microscopic image obtained by imaging in the manufacturing process of the recording / reproducing head of the hard disk device shown in FIG. (Simply referred to as “microscopic image”). In FIG. 4, the vertical direction is the y-axis, and the horizontal direction is the x-axis.

  In step S1 in FIG. 3, as the function of the edge extraction unit 512 in FIG. 2, two points at which the gradation value becomes maximum from the line profile of the gradation value in the x-axis direction at each y coordinate of each pixel of the microscope image are obtained. By extracting, the point group which comprises each edge | side of the pattern which shows the outline of the inspection object goods contained in the said microscope image is extracted. FIG. 6 shows a line profile of gradation values of each pixel on the straight line 10 in FIG.

  Next, in step S2, as a function of the approximate line acquisition unit 513, the Hough transform (described later) is performed using only the points included in the region 11 that does not include the curved portion in the left point group in FIG. The straight line 13 and the straight line 14 on the right side are obtained as approximate lines. Furthermore, only the points included in the region 12 that does not include the curved portion in the left point group are used to obtain the straight line 15 using the Hough transform, and similarly the straight line 16 for the right side as an approximate line.

  Next, in step S3, as the function of the reference point acquisition unit 514, the intersection point 17 of the straight line 13 and the straight line 15 and the intersection point 18 of the straight line 14 and the straight line 16 are obtained by calculation as the approximate lines thus obtained, For example, the midpoint 19 between them is obtained as a reference point for positioning the measurement position.

  Note that the intersection point 18 can be obtained, for example, by expressing the straight line 14 and the straight line 16 as functions relating to the xy axes and performing a function calculation.

  Next, in step S4, as a function of the measurement position acquisition unit 515, the measurement position 20 is determined from the reference point obtained in step S3 based on information registered in advance. Specifically, for example, information such as the relative coordinates of the start point of edge detection at the time of length measurement in step S5, which will be described later, with respect to the reference point, the range for searching for an edge for edge detection, the direction for searching for an edge, and the like. Based on this, the measurement position 20 is determined.

  Next, in step S5, edge detection at the measurement position 20 of the microscope image is performed as a function of the edge detection unit 516, and then, as a function of the length measurement unit 517, the detected left and right sides in FIG. The distance between the two points corresponding to the edge is calculated to perform length measurement.

  In the example of FIG. 5, in the illustrated pattern, a line profile of the gradation value of the pixel along the x direction at the measurement position 20 is obtained, and an edge is detected by one of the edge detection algorithms described later. Since the edge detected here corresponds to the outline of the pattern included in the microscope image, the width 20 at the measurement position 20 of the pattern is obtained by calculating the distance between the two points detected by the edge detection. It is done. The distance between the two points can be calculated, for example, by calculating the number of pixels between them.

  Further, in step S6, in the directory specified in the personal computer constituting the image length measuring device 50, an unmeasured image, that is, a microscope image in which the above length measurement operation is not yet performed together with steps S1 to S5. The presence or absence of is determined. The length measurement operation from step S1 to step S5 is sequentially performed on the unmeasured image.

  Finally, in step S7, as a function of the result output unit 518, the length measurement results obtained in steps S1 to S5, each straight line as an approximate line, the length measurement position, etc. are superimposed on the corresponding microscope image. Is displayed on the display of the personal computer or output as a data file.

  Next, the length measurement result of the image length measuring apparatus according to the first embodiment of the present invention described above is compared with the length measurement result using only the conventional CD-SEM, and the effect will be described.

  The 44 elements in the same semiconductor wafer were measured by CD-SEM at the measurement position 20 shown in FIG. 5, and a series of measurement work was repeated 5 times.

  On the other hand, the image data output by the CD-SEM was measured by the image length measuring device according to Example 1 of the present invention by the method described with reference to FIG.

  From these length measurement results, the variation in length measurement was repeated five times for the same element, and the histogram shown in FIG. 7 was obtained.

  In the length measurement using only the CD-SEM, the length measurement position varies due to the shift of the imaging range. As a result, the length measurement value varies greatly. In the case of the image length measurement apparatus according to the first embodiment of the present invention, Since the measurement position can be determined with high accuracy even in an image whose imaging range is shifted, variations in the measurement value can be suppressed.

  As described above, the image length measuring device according to the first embodiment of the present invention can measure a pattern at a predetermined position with high accuracy.

  Note that the imaging range for the inspection object in imaging by the CD-SEM is generally shifted in the horizontal direction in the figure, that is, in the x direction when a relatively long pattern as shown in FIGS. This is because it is difficult to increase the positioning accuracy in the vertical direction, that is, the y direction.

  Here, for example, in the case of the example of FIGS. 4 and 5, the imaging range refers to a range related to imaging in the case where the image shown in FIG. 4 is an image of a part of the inspection target product. When the imaging range is shifted in the vertical direction, even if pattern length measurement is performed at a predetermined position in the obtained image, a different portion in the inspection target product is measured, so that a length measurement error occurs.

  In the image length measuring device 50 according to the first embodiment of the present invention, the approximate line of the contour of the pattern in the microscope image is obtained as described above, and the length measurement position is determined based on the approximate line. For this reason, even if the imaging range is deviated, the length measurement position is adjusted according to the characteristics of the pattern included in the microscope image within the imaging range, so that the measurement error due to the deviation of the imaging range is minimized. It becomes possible to suppress to.

  Next, an image length measuring apparatus according to the second embodiment of the present invention will be described.

  FIG. 8 is a functional block diagram for explaining the function of the image length measuring device 50, and FIG. 9 is an operation flowchart for explaining the function of the image length measuring device 50.

  The image measuring device 50 according to the second embodiment has the same configuration as the image measuring device 50 according to the first embodiment described above, and a template matching unit 519 and a region moving unit 520 are newly added as shown in FIG. Yes.

  Here again, as in the case of the first embodiment, the microscope image of FIG.

  In step S21 in FIG. 9, by extracting two points at which the gradation value becomes maximum from the line profile of the gradation value along the x direction at each y coordinate of the pixel in the microscope image, the point group forming each side To extract.

  Next, in step S22, as a function of the template matching unit 519, template matching is performed between a so-called teacher image, which is a reference prepared in advance, and a microscope image to be measured. By performing this template matching, a deviation amount of the microscope image to be measured with respect to the teacher image is calculated.

  FIGS. 10A and 10B are diagrams for explaining the operation in this case. FIG. 10A shows a reference length measurement image 30 and FIG. 10B shows a microscopic image 35 to be measured.

  Here, the teacher image 31 in FIG. 10A is cut out from the length measurement reference image 30 having the same image size as the length measurement target microscope image 35 in FIG. 10B, and the image size is the length measurement. The size of the target microscope image 35 is half of the size.

Here, when the teacher image 31 is cut out, the coordinates (X _O1 , Y _O1 ) of the cut-out origin 32 of the teacher image 31 are stored together with the image. At the same time, the predetermined linear approximation range 33 in the measurement reference image 30 is stored as Y ≦ Y _U , and the linear approximation range 34 is stored as Y ≧ Y _L. Here, the directions of the x and y coordinates are the directions shown in the lower right part of FIG.

In template matching between the microscope image 35 to be measured and the teacher image 31, the similarity between the images is measured while changing the relative position between the teacher image 31 and the microscope image 35 to be measured. Then, a position 37 (X _O2 , Y _O2 ) on the microscope image 35 to be measured of the origin 32 of the teacher image 31 when the similarity between the images becomes maximum is obtained. In this case, the deviation amounts in the x and y directions of the microscope image 35 to be measured with respect to the teacher image 31 are obtained as differences between the two coordinates, that is, X _O2 −X _O1 and Y _O2 −Y _O1 .

Next, in step S23, only the point group included in the linear approximation ranges 38 and 39 (FIG. 10B) obtained by taking the deviation amount into account from the point group detected in step S21 is used. Each side part constituting the contour of the pattern included in the microscope image is linearly approximated. Here, the linear approximation range 38 obtained by taking the above deviation amount into account is expressed as Y ≦ Y _U + (Y _O2 −Y _O1 ), and the linear approximation range 39 is expressed as Y ≧ Y _L + (Y _O2 −Y _O1 ). It is a range.

  This is the function of the area moving unit 519. That is, according to the amount of deviation of the microscopic image obtained as a result of template matching, the linear approximation range 38, 39, that is, the target when linearly approximating each side part constituting the contour of the pattern by the function of the approximate line acquisition unit 513 By moving a rectangular window defining the range of the point cloud on the microscope image, the range is changed. As a result, the accuracy of the linear approximation can be effectively improved.

  That is, in the example of FIG. 10, as shown in the figure, the image pickup position of the pattern in the image is slightly shifted upward in FIG. That is, when the funnel-shaped portion on the upper side of the pattern is viewed, the direction in FIG. 10B appears longer in the y direction than in FIG. In other words, the funnel-shaped lower end portion is shifted downward.

  In such a case, the shift amount of the microscope image is obtained as a positive amount in the y direction. Accordingly, each rectangular window that defines a point group for performing linear approximation is moved downward. As a result, as shown in the figure, the upper straight line approximation range 38 is set slightly longer than the straight line approximation range 33 of the length measurement reference image 30, and conversely the lower straight line approximation range 39 is a straight line approximation range of the length measurement reference image 30. 34 is set slightly shorter.

  The subsequent operations in steps S24 to S28 are the same as the operations in steps S3 to S7 in the first embodiment, and a duplicate description is omitted.

  As described above, in the second embodiment, pattern matching between the teacher image and the microscopic image to be measured is performed, and a rectangular window that defines the linear approximation range of each side constituting the contour of the pattern is obtained by pattern matching. Then, the corresponding pattern is moved on the microscope image by the amount of deviation from the teacher image. As a result, it is possible to make the range of the point group that is the target for obtaining the approximate straight line of each side portion constituting the contour of the pattern with high accuracy more appropriate. As a result, it is possible to effectively improve the accuracy of obtaining the approximate curve, and it is possible to improve the accuracy of the pattern length measurement at the determined position.

  Next, an image length measuring apparatus according to a third embodiment of the present invention will be described.

  FIG. 11 is a functional block diagram for explaining the functions of the image length measuring device 50, and FIG. 12 is an operation flowchart for explaining the functions of the image length measuring device 50.

  The image length measuring device 50 according to the third embodiment has the same configuration as the image length measuring device 50 according to the first embodiment described above, and a template matching unit 521 and an optimum algorithm selecting unit 522 are newly added as shown in FIG. ing.

  In the third embodiment, it is assumed that the length measurement target image group includes three types of images A, B, and C having different pattern shapes and edge contrasts.

  In step S41 to step S44 in FIG. 12, the edge point group of each side part constituting the contour of the pattern included in the microscope image is extracted, the reference point is obtained, and the length measurement position is obtained. These operations are the same as the operations in steps S1 to S4 in the first embodiment, and a duplicate description is omitted.

  Next, in step S45, the template matching unit 521 functions as a microscope image to be measured and teacher images A, B, and C respectively corresponding to the three types A, B, and C prepared in advance. By performing template matching with each of these, the similarity between the microscopic image to be measured and these teacher images is calculated. Here, the teacher images A, B, and C are images representing the image types A, B, and C, respectively.

  Whether the image type of the microscopic image to be measured is A, B, or C depending on whether the teacher image having the greatest similarity is A, B, or C as a result of step S45. judge.

  Next, in step S46, as the function of the optimum algorithm selection unit 522, the optimum edge detection algorithm for the image type of the microscope image is selected by referring to the prepared image type / edge detection algorithm comparison table. To do.

  For example, when the image type of the microscope image is A, in the edge detection during the length measuring operation by the edge detection in step S47 corresponding to step S5 of the first embodiment, as shown in FIG. An algorithm that obtains the differential value of the gradation value of the pixel in the image, obtains the maximum value and the minimum value in the differential value line profile (differential value line profile in the figure), and obtains these as edges, that is, the differential value of the pixel value The method for obtaining the maximum and minimum values in the line profile is selected.

  On the other hand, when the image type of the microscope image is B, the maximum value in the line profile of the gradation value of the pixel at the measurement position of the microscope image in FIG. 9 (“gradation value line profile” in FIG. 9) is edged. Is selected, that is, a method for obtaining a maximum value in a line profile of pixel gradation values.

  When the image type of the microscope image is C, in FIG. 9, in the line profile of the gradation value of the pixel at the measurement position of the microscope image, the point where the gradation value exceeds a certain threshold is indicated by the gradation value line. An algorithm for obtaining the two points as the maximum value in the profile by searching toward both sides starting from the start point, that is, an algorithm for obtaining these points as edges, that is, a method for obtaining the threshold value in the line profile of the gradation value of the pixel is selected.

  In step S47, edge detection at the length measurement position of the microscope image is performed by the algorithm selected in this way, and the width of the pattern at the position is measured. The operations in steps S47 to S49 are the same as the operations in steps S5 to S7 in the first embodiment, and a duplicate description is omitted.

  In a microscope image obtained by a scanning electron microscope, the edge contrast at the time of imaging may vary greatly depending on conditions such as the uneven shape and material on the inspection object. In order to accurately measure the length from the microscope image obtained under such conditions, it is necessary to apply an optimum edge detection algorithm according to each condition.

  Conventionally, the operator selects the optimum edge detection algorithm by himself / herself. On the other hand, according to the image length measurement device of the third embodiment, as a preparation before length measurement, a comparison table between the teacher image corresponding to the image type and the corresponding optimum edge detection algorithm is created in advance. In the process, the optimum edge detection algorithm is automatically selected and applied according to the result of template matching. Therefore, a significant reduction in measuring man-hours can be realized.

  The edge detection algorithm, that is, the edge detection method will be described later with reference to FIG.

  Next, an image length measuring apparatus according to a fourth embodiment of the present invention will be described.

  FIG. 14 is a functional block diagram for explaining the functions of the image length measuring device 50, and FIG. 15 is an operation flowchart for explaining the functions of the image length measuring device 50.

  The image length measuring device 50 according to the third embodiment has the same configuration as the image length measuring device 50 according to the first embodiment, and a template matching unit 523 and a similarity determination unit 524 are newly added as shown in FIG. ing.

  In the fifth embodiment, first, in step S50 in FIG. 15, as a function of the template matching unit 523, both images are obtained by performing template matching with a teacher image prepared in advance for a microscopic image to be measured. Align between them. In this way, after correcting the deviation between the two, the similarity between the teacher image and the microscope image to be measured is calculated.

  Next, in step S51, as a function of the similarity determination unit 524, reference is made to a threshold value of similarity prepared in advance. If the obtained similarity is larger than the threshold, subsequent steps S52 to S56 are performed. Perform the length measurement operation. The length measuring operations in steps S52 to S56 are the same as those in steps S1 to S5 in the first embodiment, and redundant description is omitted.

  On the other hand, if the degree of similarity is smaller than the threshold value (“low” in step S51), it is determined that the microscope image is a different type of image that does not need to be measured, and the process moves to step S57 as it is and another microscope image is obtained. Perform the operation to perform the length measurement operation.

  In automatic measurement of semiconductor wafers with CD-SEM, focusing is performed automatically in addition to visual field movement, but images are taken out of focus due to differences in the surface condition of the inspection object. Cases can arise. In addition, there may be a case where there is some trouble during the manufacturing process, and the element part to be measured is partially or completely destroyed.

  As described above, there is a possibility that the microscopic image related to the image picked up by the CD-SEM includes an image that is not suitable for the length measurement processing as described above based on the image, or an image that is not necessary for it. . According to the image length measurement apparatus of the above-described fourth embodiment, when the measurement target image group includes such unsuitable or unnecessary images, they are automatically excluded from the length measurement process. As a result, it is possible to omit useless length measurement processing, and therefore it is possible to effectively reduce the total length measurement man-hour when processing a number of microscope images to be measured.

  Next, an image length measuring apparatus according to a fifth embodiment of the present invention will be described.

  In the fifth embodiment, the computer 110 itself that controls the CD-SEM 100 functions as an image measuring device.

  FIG. 16 is a block diagram for explaining the function of the image length measuring device 110, and FIG. 17 is an operation flowchart for explaining the function of the image length measuring device 110.

  The image length measuring device 110 according to the fifth embodiment has the same configuration as that of the image length measuring device 50 according to the above-described first embodiment or the like in a portion that performs the length measuring operation. In this case, as shown in FIG. 16, the personal computer 110 itself that controls the CD-SEM 100 functions as an image length measuring device as described above.

  First, in step S60 of FIG. 17, the computer 110 itself is monitored. Then, for example, in step S61, the number of files existing in the directory in which images in the computer 110 are stored is checked. If the number of files has increased ("Yes" in step S61), the increase in step S62. A length measurement operation is performed on the microscope image, and the length measurement result is output in step S63. The microscopic image associated with the increase in the directory can be determined by, for example, examining the time stamp of the corresponding image file.

  The length measuring operation performed in step S62 is the length measuring operation in each of the above-described first to fourth embodiments, that is, among the length measuring operations described above together with the respective operation flowcharts of FIGS. 3, 9, 12, and 15. Any of these may be used. Here, the process of step S62 includes a series of operations from extraction of a point group constituting each side part constituting the contour of the pattern included in the microscope image to measurement of the pattern dimension by edge detection. It is desirable to have an automatic selection function of an edge detection algorithm as in the third embodiment described above with reference to FIG.

  In the conventional image length measuring apparatus, it is necessary for the operator to determine and designate a microscope image or a group of images to be subjected to the length measuring work. As a result, when the number of images is enormous, a situation may be assumed in which the operation of selecting an image occupies most of the man-hours of the image measurement operation. According to the image length measurement apparatus of the fifth embodiment of the present invention, the length measurement operation is automatically performed almost completely without human intervention, and the operator can obtain the length measurement result without any particular delay time. I can know.

  In other words, according to the fifth embodiment, when a CD-SEM 100 acquires a microscope image of a product to be inspected and saves it in a predetermined directory such as a hard disk device of the computer 110, this is automatically detected and the new image is obtained. The length measurement operation for the microscopic image for storage is automatically started. Therefore, the operator simply performs an operation of saving the microscope image in the predetermined directory after acquiring the microscope image, and thereafter, the length measurement operation for the microscope image is automatically executed by the function of the image measuring device, The operator can obtain the result.

  The embodiments of the present invention have been described above. However, the embodiments of the present invention are not limited to these configurations, conditions, and the like, and various modifications can be made. For example, in each of the above embodiments, each side portion constituting the contour of the pattern included in the microscope image is approximated by a straight line, but may be approximated by a curve (described later with reference to FIG. 19).

  In the second embodiment, the rectangular window that defines the range of the point group to be linearly approximated is moved by the amount of deviation obtained by pattern matching. You may perform in the step which detects the point group which makes each edge part which comprises an outline.

  In the third embodiment, the optimum edge detection algorithm is selected and applied by determining the image type by template matching. On the other hand, by preparing a comparison table of image type, optimal edge detection algorithm, and measurement position in advance, it is possible to obtain the measurement position along with the optimal edge detection algorithm according to the determination result of the image type. it can. And it is good also as a structure which performs length measurement operation | movement by the said optimal edge detection algorithm in the said length measurement position. Furthermore, the process of performing template matching can be changed. For example, template matching is performed at the stage before extracting the edge point group of each side that constitutes the contour of the pattern, and the optimum edge detection algorithm is selected, and the selected algorithm is used for each side constituting the contour of the pattern. You may apply as an algorithm which extracts the edge point group of a part.

  FIG. 18 is a block diagram showing the configuration of the computer 50 or 110 constituting each embodiment of the present invention.

  As shown in FIG. 18, the computer 50 or 110 includes a CPU 51 for executing various operations by executing instructions constituting a given program, a keyboard, a mouse, etc. An operation unit 52 for inputting, a display unit 53 composed of a CRT, a liquid crystal display, etc., which displays processing progress and processing results by the CPU 51 to the user, a program executed by the CPU 54, data, etc. composed of ROM, RAM, etc. A memory 54 used for storage or as a work area, a hard disk device 55 for storing programs, data, and the like; a CD-ROM drive 56 for loading programs and data from outside via a CD-ROM 57; Or an external server via a communication network 59 such as the Internet or a LAN. And a modem 58 for download, such as a program.

  The computers 50 and 110 use the CD-ROM 57 as an intermediary or the communication network 59 as an intermediary to perform processing such as a length measurement operation performed by the image length measurement device of each of the above-described embodiments, that is, FIGS. , Loads or downloads a program comprising instructions for causing the CPU 51 to execute the operations shown in the respective operation flowcharts of FIG. Then, this is installed in the hard disk device 55, loaded into the memory 54 as appropriate, and executed by the CPU 51. As a result, an image measuring device is realized by the computers 50 and 110.

  In the following, an example other than the straight line approximation in the case where the approximate line is obtained from the sides constituting the contour of the pattern included in the microscope image in the approximate line acquisition unit 513 in each of the above embodiments will be described.

  Here, an example of approximation by a circle (or ellipse) is shown.

When it is desired to measure the width w of the arrow portion in the pattern as shown in FIG. 19A, each curved portion surrounded by a broken line in FIG. 19B is approximated by an arc and the center (x _c1 , y _c1 ) And (x _c2 , y _c2 ), the midpoint (x ₀ , y ₀ ) = ((x _c1 + x _c2 ) / 2, (y _c1 + y _c2 ) / 2) is taken as the reference point.

In this example, a position at a distance d from the reference point is defined as a measurement position, and the width w is measured by performing edge detection along the x direction at a position where the y coordinate is y ₀ + d.

  Next, the Hough transform for obtaining the approximate line of the side constituting the outline of the pattern included in the microscope image in the approximate line acquisition unit 513 will be described.

  Hough conversion is described, for example, on pages 204 to 206 of “Computer Image Processing” edited by Hideyuki Tamura (Ohm Corp., 2002).

That is, if the shape of the line to be detected is predetermined and the shape can be expressed by an algebraic equation, the Hough transform (which maps the feature points (edge points, etc.) in the image to the parameter space representing the shape of the line ( Hough
(Transform) is considered to be effective. Here, the most standard Hough transform for line detection will be described.

First, as an algebraic equation representing a straight line,

ρ = x cos θ + ysin θ. . . (B-1)

Is used.

In the above equation, ρ is a length of a perpendicular line drawn straight from the coordinate origin, and θ is a parameter representing an angle between the perpendicular line and the x axis. Using this algebraic equation, a straight line ρ ₀ = x ₀ cos θ + y ₀ sin θ in the xy image space shown in FIG. 20A is expressed as one point in the ρ-θ parameter space shown in FIG. Is done.

An arbitrary straight line passing through the feature points (x ₀ , y ₀ ) in the image is

ρ = x ₀ cos θ + y ₀ sin θ. . . (B-2)

The straight line group diagram 20 (c) satisfying this equation forms a locus as shown in FIG. 20 (d) in the parameter space.

In other words, this means that the point (x ₀ , y ₀ ) in the image space is mapped to the locus in the parameter space represented by the equation (B-2).

  Based on the relationship between the image space and the parameter space, the Hough transform performs straight line detection as follows.

  1) A two-dimensional array representing a parameter space is prepared, and all of its values are initialized to zero.

  2) An equation obtained by substituting the coordinate values of each feature point in the image into x and y in the equation (B-1) is regarded as an equation relating to ρ and θ, and a locus represented by the equation is drawn in the parameter space (FIG. 20). (See (e)).

  To draw a locus, the array element through which the locus passes is obtained by calculating the value of ρ satisfying the equation while increasing θ by a constant interval Δθ, and the locus is overwritten by incrementing the value by 1 (voting). .

  3) After drawing trajectories corresponding to all feature points, a position where a large number of trajectories are concentrated, that is, an element having a large maximum value in an array representing the parameter space is obtained.

Let the parameters represented by such elements be (ρ ^* , θ ^* ). This parameter represents a linear ρ ^* = xcosθ ^* + ysinθ ^* in the image, the straight line passing through a large number of feature points will have been detected (see FIG. 20 (e), (f) ).

  Hough transform is said to be able to detect a straight line well even if feature points in the image are not continuous, and to be more resistant to noise than edge tracking. As described above, the Hough transform has excellent characteristics, and since the first proposed 1962, many improvements and extensions have been made.

  The Hough transform can be used to detect circles and ellipses in addition to straight lines. However, in the case of a circle, three parameters are required, and in the case of an ellipse, five parameters are required. In the simple method, it is necessary to prepare a three-dimensional array and a five-dimensional array, respectively. It is said that there is a problem.

  Various ideas have been considered for this problem. Further, there is a generalized Hough transform as a method for detecting a two-dimensional figure that cannot be expressed by a simple algebraic equation such as a polygon. Furthermore, the concept of Hough transform can also be used for detection and recognition of a three-dimensional plane or object.

  Next, a method for edge detection by the edge extraction unit 512 or the edge detection unit 516 will be described.

  For the edge detection method, for example, the roadmap on the homepage of the Japan Semiconductor Manufacturing Equipment Association (2003) (URL: http://www.seaj.or.jp/rdmp/indx_f.htm, October 2006) (As of the 12th) is shown in the fifth edition “Measurement”.

  That is, there are various systems shown in FIG. 21 that are actually applied to the CD-SEM as using the line profile of the gradation value of the pixel of the microscope image. In general, there are many other methods, and as described with reference to FIG. 13, there is also a method using a line profile of differential values. In FIG. 21, L indicates a length measurement value, and A indicates an actual pattern width.

The present invention can take the configurations described in the following supplementary notes.
(Appendix 1)
Approximate line acquisition means for obtaining an approximate line from a point group constituting a given pattern;
A reference point obtaining means for obtaining a reference point by a predetermined calculation from the approximate line;
A measurement position acquisition means for obtaining a measurement position from the reference point by a predetermined calculation;
A pattern dimension measuring apparatus comprising dimension measuring means for measuring a dimension of a predetermined portion of the pattern at the length measurement position.
(Appendix 2)
The pattern dimension measuring apparatus according to appendix 1, wherein the approximate line acquisition means includes means for approximating the point group with one or a plurality of straight lines.
(Appendix 3)
The pattern dimension measuring apparatus according to supplementary note 1, wherein the approximate line acquisition means includes means for approximating the point group with one or a plurality of predetermined curves.
(Appendix 4)
The pattern dimension measuring apparatus according to appendix 3, wherein the curve is an arc.
(Appendix 5)
The pattern dimension measuring apparatus according to claim 1, wherein the approximate line acquisition unit includes a unit that approximates the point group with a straight line or a curve represented by a predetermined function.
(Appendix 6)
The pattern dimension measuring apparatus according to any one of supplementary notes 1 to 5, wherein the approximate line acquisition unit is configured to obtain an approximate line using a Hough transform.
(Appendix 7)
The pattern dimension measuring device according to any one of appendices 1 to 6, wherein the reference point acquisition stage is configured to obtain a reference point based on an intersection of the approximate lines obtained in the approximate line acquisition stage. .
(Appendix 8)
The approximate line acquisition means is intended for a point group belonging to a predetermined region among the point groups constituting the given pattern,
The pattern dimension measurement according to any one of appendices 1 to 7, further comprising means for moving the predetermined area based on a result of pattern matching between the given pattern and a predetermined reference pattern. apparatus.
(Appendix 9)
The dimension measuring means uses an edge detection method for determining the position of the edge of the predetermined part when measuring the dimension of the predetermined part of the pattern according to a change mode of the pixel value constituting the pattern according to a change mode. The pattern dimension measuring apparatus according to any one of supplementary notes 1 to 8, wherein the pattern size measuring apparatus is configured to select from the plurality of methods.
(Appendix 10)
The predetermined plurality of methods include a method of obtaining a maximum value and a minimum value in a line profile of a differential value of a pixel value, a method of obtaining a maximum value in a line profile of a pixel gradation value, and a line profile of a pixel gradation value The pattern dimension measuring apparatus according to appendix 9, wherein the pattern dimension measuring apparatus includes a method for obtaining the threshold value.
(Appendix 11)
Further, before the operation for obtaining the approximate line by the approximate line acquisition unit is executed, the given pattern is subjected to pattern matching with a predetermined reference pattern, and the correlation with the predetermined reference pattern is smaller than a predetermined value. In this case, the pattern dimension according to any one of appendices 1 to 10, wherein the approximate line acquisition unit, the reference point acquisition unit, the measurement position acquisition unit, and the dimension measurement unit do not execute a series of operations. measuring device.
(Appendix 12)
A series of operations by the approximate line acquisition unit, the reference point acquisition unit, the length measurement position acquisition unit, and the dimension measurement unit are configured to be automatically executed by computer control,
A storage detection unit that detects storage of an image by an image storage unit that stores an image obtained by an imaging unit that obtains an image of the given pattern by imaging;
A configuration in which a series of operations by the approximate line acquisition unit, the reference point acquisition unit, the length measurement position acquisition unit, and the dimension measurement unit are automatically started upon detection of image storage by the storage detection unit. The pattern dimension measuring apparatus according to any one of appendices 1 to 11,
(Appendix 13)
An approximate line acquisition stage for obtaining an approximate line from a point group constituting a given pattern;
A reference point obtaining step for obtaining a reference point from the approximate line;
A measurement position acquisition step for obtaining a measurement position from the reference point;
A pattern dimension measuring method comprising a dimension measuring step of measuring a dimension of a predetermined portion of the pattern at the length measurement position.
(Appendix 14)
The approximate line acquisition step is performed for a point group belonging to a predetermined region among the point groups constituting the given pattern,
14. The method according to appendix 13, wherein the method includes a step of moving the predetermined area based on a result of pattern matching between the given pattern and a predetermined reference pattern.
(Appendix 15)
In the dimension measurement step, an edge detection method for obtaining the position of the edge of the part when measuring the dimension of the predetermined part of the pattern is determined according to a change state of the pixel value constituting the pattern in the part. 15. The method according to appendix 13 or 14, wherein the method is selected from the plurality of methods.
(Appendix 16)
Furthermore, before the approximate line acquisition step, pattern matching is performed with a predetermined reference pattern for the given pattern, and when the correlation with the predetermined reference pattern is smaller than a predetermined value, the series of approximate lines The method according to any one of supplementary notes 13 to 15, wherein the acquisition stage, the reference point acquisition stage, the measurement position acquisition stage, and the dimension measurement stage are not executed.
(Appendix 17)
An approximate line acquisition stage for obtaining an approximate line from a point group constituting a given pattern;
A reference point obtaining step for obtaining a reference point from the approximate line;
A measurement position acquisition step for obtaining a measurement position from the reference point;
A program comprising instructions for causing a computer to execute a dimension measuring step of measuring a dimension of a predetermined portion of the pattern at the length measurement position.
(Appendix 18)
The approximate line acquisition step is performed for a point group belonging to a predetermined region among the point groups constituting the given pattern,
18. The program according to appendix 17, wherein the program includes a command for causing a computer to execute a step of moving the predetermined region based on a result of pattern matching between the given pattern and a predetermined reference pattern.
(Appendix 19)
In the dimension measurement step, an edge detection method for obtaining the position of the edge of the part when measuring the dimension of the predetermined part of the pattern is determined according to a change state of the pixel value constituting the pattern in the part. The program according to appendix 17 or 18, wherein the program is selected from the plurality of methods.
(Appendix 20)
Furthermore, before the approximate line acquisition step, pattern matching is performed with a predetermined reference pattern for the given pattern, and when the correlation with the predetermined reference pattern is smaller than a predetermined value, the series of approximate lines The program according to any one of supplementary notes 17 to 19, wherein the acquisition stage, the reference point acquisition stage, the length measurement position acquisition stage, and the dimension measurement stage are not executed.

  The application example of the present invention is not limited to the pattern length measurement for the scanning electron microscope image in the manufacturing process of the semiconductor device and the hard disk device as described above. For example, the present invention is applicable to all industries that handle microscopic images. The imaging apparatus for obtaining the measurement target image is not limited to the scanning electron microscope, but may be an optical microscope, a CCD camera, a transmission electron microscope, or the like. The present invention measures a pattern included in an image captured by these. It is also applicable to the head.

It is a block diagram for demonstrating the image length measuring apparatus by the Example of this invention. It is a functional block diagram for demonstrating the function of the image length measuring apparatus by Example 1 of this invention. It is an operation | movement flowchart for demonstrating operation | movement of the image length measuring apparatus by Example 1 of this invention. It is a figure which shows the image of an example of the length measuring object of the image length measuring apparatus by the Example of this invention. It is a figure for demonstrating the length measuring method by the image length measuring apparatus by Example 1 of this invention. FIG. 5 is a diagram showing an example of a line profile of gradation values in the image shown in FIG. 4. It is a figure for demonstrating the function of the image length measuring apparatus by Example 1 of this invention. It is a functional block diagram for demonstrating the function of the image length measuring apparatus by Example 2 of this invention. It is an operation | movement flowchart for demonstrating operation | movement of the image length measuring apparatus by Example 2 of this invention. It is a figure for demonstrating the positioning procedure of the length measurement position in the image length measuring apparatus by Example 2 of this invention. It is a functional block diagram for demonstrating the function of the image length measuring apparatus by Example 3 of this invention. It is an operation | movement flowchart for demonstrating operation | movement of the image length measuring apparatus by Example 3 of this invention. It is a figure for demonstrating the edge detection algorithm in the image length measuring apparatus by Example 3 of this invention. It is a functional block diagram for demonstrating the function of the image length measuring apparatus by Example 4 of this invention. It is an operation | movement flowchart for demonstrating operation | movement of the image length measuring apparatus by Example 4 of this invention. It is a block diagram for demonstrating the function of the image length measuring apparatus by Example 5 of this invention. It is an operation | movement flowchart for demonstrating operation | movement of the image length measuring apparatus by Example 5 of this invention. It is a block diagram for demonstrating the structure of the said computer in the case of implement | achieving the image length measuring apparatus by the Example of this invention with a computer. It is a figure for demonstrating the method of approximating the outline of the given pattern with a circular arc applicable in the image length measuring apparatus by the Example of this invention. It is a figure for demonstrating Hough conversion. It is a figure for demonstrating the edge detection method.

Explanation of symbols

10 Boundary lines 11, 33, and 38 of the approximate side of the oblique side Approximate range of the oblique side 12, 34, and 39 Approximate range of the vertical side 13 and 14 Approximate line 15 and 16 of the oblique side Approximate line 17 and 18 of the vertical side Intersection of straight line and vertical approximate straight line 19 Measurement reference point 20 Measurement position example 30 Measurement reference image 31 Teacher image cutout range 32 Teacher image cutout origin 35 Microscope image 36 subject to length measurement 37 Area detected by template matching 37 Template Origin 50 of region detected by matching Image length measuring device 100 CD-SEM
511 Image acquisition unit 512 Edge extraction unit 513 Approximate line acquisition unit 514 Reference point acquisition unit 515 Measurement position acquisition unit 516 Edge detection unit 517 Length measurement unit 518 Result output unit 519, 521, 523 Template matching unit 520 Region movement unit 522 Optimum Algorithm selection unit 524 Similarity determination unit

Claims (7)

Approximate line acquisition means for obtaining an approximate line from a point group constituting a given pattern;
A reference point obtaining means for obtaining a reference point from the approximate line;
A measurement position acquisition means for obtaining a measurement position from the reference point;
A pattern dimension measuring apparatus comprising dimension measuring means for measuring a dimension of a predetermined portion of the pattern at the length measurement position. The approximate line acquisition means is intended for a point group belonging to a predetermined region among the point groups constituting the given pattern,
The pattern dimension measuring apparatus according to claim 1, further comprising means for moving the predetermined area based on a result of pattern matching between the given pattern and a predetermined reference pattern.   The dimension measuring means uses an edge detection method for determining the position of the edge of the predetermined part when measuring the dimension of the predetermined part of the pattern according to a change mode of the pixel value constituting the pattern according to a change mode. The pattern dimension measuring apparatus according to claim 1, wherein the pattern dimension measuring apparatus is configured to be selected from a plurality of methods.   Further, before the operation for obtaining the approximate line by the approximate line acquisition unit is executed, the given pattern is subjected to pattern matching with a predetermined reference pattern, and the correlation with the predetermined reference pattern is smaller than a predetermined value. In this case, the approximate line acquisition unit, the reference point acquisition unit, the measurement position acquisition unit, and the dimension measurement unit are not configured to execute a series of operations. Pattern dimension measuring device. A series of operations by the approximate line acquisition unit, the reference point acquisition unit, the length measurement position acquisition unit, and the dimension measurement unit are automatically executed by computer control,
A storage detection unit that detects storage of an image by an image storage unit that stores an image obtained by an imaging unit that obtains an image of the given pattern by imaging;
A series of operations by the approximate line acquisition unit, the reference point acquisition unit, the length measurement position acquisition unit, and the dimension measurement unit are automatically performed when the storage detection unit detects the image storage by the image storage unit. The pattern dimension measuring apparatus according to claim 1, wherein the pattern dimension measuring apparatus is configured to be started. An approximate line acquisition stage for obtaining an approximate line from a point group constituting a given pattern;
A reference point obtaining step for obtaining a reference point from the approximate line;
A measurement position acquisition step for obtaining a measurement position from the reference point;
A pattern dimension measuring method comprising a dimension measuring step of measuring a dimension of a predetermined portion of the pattern at the length measurement position. An approximate line acquisition stage for obtaining an approximate line from a point group constituting a given pattern;
A reference point obtaining step for obtaining a reference point from the approximate line;
A measurement position acquisition step for obtaining a measurement position from the reference point;
A program comprising instructions for causing a computer to execute a dimension measuring step of measuring a dimension of a predetermined portion of the pattern at the length measurement position.