JP3900368B1 - LAYER SECTION MEASURING SYSTEM AND METHOD, AND COMPUTER PROGRAM - Google Patents

Abstract

In a moving state analysis system, a moving state or a changing state of an image can be analyzed with high accuracy.
A dynamic state analysis system (100) includes a difference calculation unit (100) for calculating a difference between at least two images arranged at a predetermined interval on a time axis or a quantity axis correlated with time in a pixel unit. 111) and a statistical difference value such as an average value or a maximum value of the calculated differences for a plurality of pixels forming an array intersecting at least in one direction on the image, and each point on the coordinate axis along the one direction. Difference statistical value calculation means (112) for calculating in association with each other, and an analysis for generating an analysis image by placing the calculated difference statistical value on a plane stretched between a coordinate axis and a time axis or a quantity axis Image generating means (113).
[Selection] Figure 1

Description

The present invention, for example, a layer cross-sectional measurement system and method for analyzing a change state in an image changing with respect to a quantity axis correlated with a distance on a scanning line set in a direction intersecting the surface of a laminate to be analyzed, And a computer program.

An image analysis system including this type of layer cross-section measurement system uses a “spatio-temporal image method” which is one of the traditional numerical methods. When using the spatio-temporal image method, the density value image of the two-dimensional plane stretched by the horizontal coordinate axis and the vertical coordinate axis is arranged on the time axis as the third axis on the time axis in the horizontal or vertical direction of the spatio-temporal image. A moving state is described by numerical values obtained by pattern analysis of a spatiotemporal cross-sectional image obtained by cutting in a direction along the direction. In particular, the “time-series difference image method”, which is a kind of such a method, takes a difference between two images in the time series, extracts only a changed portion in the image, and performs pattern analysis on the difference image. Therefore, the dynamic state is quantified.

Conventionally, as this type of image analysis system, for example, in the technical field of performing autopilot and steering support in a vehicle, the object to be digitized in an image captured by an in-vehicle camera is identified by pattern recognition technology, and then the target A system for digitizing the movement of an object by the above-described digitization method has been developed (see Patent Document 1).

In addition, such an image analysis system, for example, measures the layer thickness of a phosphor in a plasma display, measurement of the layer thickness of a polarizing plate composed of laminated films, measurement of the layer thickness of the phosphor and interlayer gap, and the layer cross section of the laminated structure. It also functions as the main part of the measurement system. Here, a plurality of surface images or optical interface images captured at a constant scanning speed with a direction perpendicular to the surface of the measurement object (in other words, a depth direction) as a scanning direction are continuous images constituting the moving image as described above. Therefore, if the dynamic state is analyzed, it is possible to analyze the change state of each layer in a plurality of surface images or optical interface images instead of the actual dynamic state.

JP 2001-175845 A

However, according to the background art described above, there is a technical problem that an S / N (Signal to Noise) ratio in an image or image data is not necessarily high enough. For example, according to the above-described time-series difference image method, the noise component in the difference or the difference value is relatively large. Therefore, the noise component becomes larger or smaller, which is an obstacle to pattern analysis of the difference image .

  In addition, according to the layer cross-sectional measurement system described above, an interference profile is usually used when acquiring an image, and the sensitivity is generally poor or the signal is weak. The noise is remarkable. As a result, there is a technical problem that it is difficult or practically impossible to accurately measure the layer thickness and the interlayer gap.

The present invention has been made in view of the above-mentioned problems, and provides, for example, a layer cross-sectional measurement system and method capable of analyzing a change state of an image with high accuracy, and a computer program that causes a computer to function as such a system. The task is to do.

In order to solve the above problems, the layer cross-sectional measurement system of the present invention has at least two sheets arranged at a predetermined interval on a quantity axis that is correlated with the distance on the scanning line set in the direction intersecting the surface of the laminate . Difference calculation means for calculating the difference of the image in units of pixels, and the average value of the calculated difference and the maximum value of the calculated difference for a plurality of pixels forming an array intersecting in at least one direction on the image And difference statistical value calculation means for calculating a difference statistical value corresponding to each point on the coordinate axis along the one direction, which is at least one of the number of pixels in which the calculated difference is equal to or greater than a threshold value. And an analysis image generation means for generating an analysis image by placing the calculated difference statistical value on a surface stretched between the coordinate axis and the quantity axis .

According to PBN meter measuring system of the present invention, the distance between at least two images arranged at predetermined intervals on the quantity axis is correlated multiple images are subject to the layer cross-section analysis. Such at least two images are, for example, a plurality of layers arranged along a scanning line for layer cross-sectional analysis (typically, a scanning line set in a direction perpendicular to the surface of the stacked structure, that is, a depth direction). A plurality of surface views or optical interface diagrams (typically laminated structures) obtained by performing continuous or intermittent scanning at a constant speed along such scanning lines. 2 is an optical interface diagram at each depth in FIG. 2). FIG.

At the time of analysis, first, for example, by a difference calculation means including a processor, a memory, etc., for example, a transparent multilayer structure, a direction intersecting the surface of the laminate, for example, perpendicularly (that is, along the depth direction in the laminate). The difference between at least two images arranged at a predetermined interval on the quantity axis correlated with the distance on the set scanning line is calculated in units of pixels. Here, “difference between at least two images” means that each color of a single color or RGB system or CMY system at the same pixel (that is, a pixel at the same position or the same address) between at least two images. Means a difference or difference value of parameters that define an image, such as density information, luminance information, saturation information, color information, and tone information. At this time, “difference between two images” or “difference between only two images” means a specific pixel in one image and the same specific pixel in another image before or after, It is a difference or a difference value for a parameter to be calculated. The “difference of three or more images” is an average of parameters to be calculated with respect to a specific pixel in one image and the same specific pixel in a plurality of other images before or after. It is a difference or a difference value as a deviation from a reference value such as a value (that is, a reference value based on a parameter of a specific pixel in three or more images). In a simple case, the density difference in black and white display between two images is calculated. Moreover, the image from which such a difference is calculated may be a digital image or an analog image. Furthermore, “difference” or “difference value” according to the present invention is a concept including “differentiation” or “differentiation value” which is a minute change with respect to the quantity axis. As described above, the “difference” according to the present invention means, in a narrow sense, a simple difference between parameters to be calculated for the same pixel between two images, and in a broad sense, three or more images. It may also mean a difference or deviation from a reference value of a parameter to be calculated for the same pixel.

  Subsequently, an array intersecting in one direction (for example, orthogonal) at each point of the coordinate axis set along the one direction on the image by the difference statistic value calculation means including, for example, a processor and a memory. Difference statistical values of a plurality of pixels are calculated. Here, “one direction” refers to, for example, an arbitrary direction such as a horizontal image, a vertical direction, both horizontal and vertical directions, an oblique direction, or any two of a single image, such as a bitmap image or a raster image. The direction is preferably set according to the property of the moving image to be analyzed by the dynamic state analysis system. Such an array of pixels is typically an array of pixels constituting the same image, that is, an image at the same time or an image acquired at the same opportunity, and a one-dimensional array ( That is, it may be a single pixel column) or a two-dimensional array that includes such a one-dimensional array or spreads in a plane within the same image (that is, a plurality of pixel columns or a group of pixels in a predetermined plane area). May be. In other words, the pixel array is located at least on a line segment extending in the direction intersecting the one direction on one image.

However, such an arrangement of pixels may extend over the plurality of images when a plurality of images are arranged along the quantity axis . That is, such an array of pixels is typically a one-dimensional array or a two-dimensional array on the same image, but a one-dimensional array or a two-dimensional array over a plurality of images. Alternatively, a three-dimensional array may be used. The “difference statistical value” according to the present invention is at least one of the average difference value, the maximum difference value, and the number of pixels in which the calculated difference is equal to or greater than a threshold value. It is also possible to use an existing statistical value known as a statistical value other than these average values related to the difference as the differential statistical value here.

Subsequently, an analysis image is generated by placing the calculated difference statistical value on a plane stretched between the coordinate axis and the quantity axis by an analysis image generation unit including a processor, a memory, and the like. Is done. Here, for example, an analysis image is generated at each point on the two-dimensional plane of the coordinate axis and the quantity axis, in which the difference statistical value is indicated by a different color density or color depending on the value. Mathematically, a corresponding difference statistic is assigned for each point on a two-dimensional plane with the coordinate axis and quantity axis as two orthogonal axes, in other words, perpendicular to this two-dimensional plane. A quantity axis indicating a difference statistic value is set as the third axis, and a color, density, or the like that changes according to the difference statistic value at each point on the two-dimensional plane is assigned to each point. Patterns such as color and density are revealed. Here, in particular, for example, based on a plurality of surface views or optical interface views related to the laminated structure, a coordinate axis (that is, X axis or Y axis) and an axis extending in the depth direction as a quantity axis (that is, Z axis) A cross-sectional view stretched by is generated as an analysis image. The analysis image generated in this way is subjected to pattern analysis, for example, automatically by pattern analysis means for pattern analysis of this analysis image or by visual analysis by an analyst, so that the original image (that is, a series of still images) And a change state in a plurality of surface diagrams or optical interface diagrams related to one laminated structure) can be analyzed.

  Among such a series of processes, when calculating the difference statistic value by the difference statistic value calculating means, a random noise component or a relatively highly random noise component in the image or image data is converted into the image or image data. It is possible to relatively weaken a signal component having a relatively low randomness. That is, the “difference statistical value” according to the present invention, which is the number of pixels whose average value, maximum value, or difference is equal to or greater than the threshold value, has a higher S / N ratio than a simple difference value or a simple differential value. In other words, even if the signal component in the image or the image data is weak, the signal component is relatively strengthened at the stage of calculating the difference statistical value. In other words, the “difference statistical value” according to the present invention is not limited to the average value, the maximum value, or the number of pixels in which the difference is equal to or greater than the threshold value, and the S / N ratio is more or less compared to the difference itself. It is sufficient that the difference is a value obtained by statistical processing so that the signal component becomes higher or the signal component is relatively strengthened.

In particular, in the present invention, by using the “average” of the difference as an example of the difference statistical value, the S / N ratio can be improved as described above, and a uniform but weak change state in the field of view can be extracted as a significant signal. .

By using the “maximum value” of the difference as another example of the difference statistical value, it is possible to extract a change state that occurs locally in the field of view as a signal. In this case, if the “average” of the differences is used, the extracted signal is weakened because the change state is local.

Furthermore, as another example of the difference statistic value, by using the “number of pixels equal to or greater than the threshold value” relating to the difference, a signal having a change state having an extremely large difference and a signal having a relatively small change state having a difference greater than or equal to the threshold are obtained. Since these signals can be treated as the same weight, it is possible to prevent a signal with a relatively small change state from being hidden by an extremely large change state signal. In this case, if the above-mentioned “average” or “maximum value” of the differences is used, a relatively weak change state signal is hidden by an extremely large change state signal.

As a result of the above, in the analysis image generated by the analysis image generation means, when directly analyzing the density of each pixel on the original image, in particular, the pixel-by-pixel difference between the two images. Compared to the above, even if the noise component in the image or the image data is large, the change state of the image can be analyzed with higher accuracy.

Therefore, according to such a layer cross-section measurement system, it is possible to analyze a layer cross-section with high accuracy, more specifically, it is possible to measure a layer thickness and an interlayer gap with high accuracy, and to accurately measure the temperature of the layer thickness. Change can be measured.

In one aspect of the layer cross-section measurement system of the present invention, pattern analysis means for performing pattern analysis on the generated analysis image is further provided.

According to this aspect, for example, the analysis image generated by the analysis image generation unit can be automatically analyzed by the pattern analysis unit including a processor, a memory, and the like, which is advantageous in practice. For example, layer thickness measurement, interlayer gap measurement, layer thickness change measurement, and the like can be automatically executed on the analysis image by the pattern analysis means.

In another aspect of the layer cross-section measurement system of the present invention, there is provided an imaging unit that generates a plurality of images by receiving electromagnetic waves including visible light from the laminated body at a plurality of measurement points arranged on the scanning line. The difference calculation means further calculates a difference between the at least two images included in the plurality of images.

According to this aspect, for example, by an imaging unit including an optical microscope using interference with reference light, for example, in a direction crossing the surface of the laminate, which is typically a transparent multilayer structure , for example, perpendicular Electromagnetic waves are received from the laminated body at a plurality of measurement points arranged on the scanning line set in (i.e., along the depth direction in the laminated body). Such reception or reception of electromagnetic waves such as reception of visible light is typically performed continuously by the imaging means moving its focal position or imaging position at a constant speed or regularly along the scanning line. Or it is performed intermittently. By such imaging with scanning, a plurality of images arranged at predetermined intervals or constant intervals on the quantity axis correlated with the distance are generated. Then, the difference calculation means calculates the difference between at least two images included in the plurality of images for each pixel, and further calculates the difference statistical value calculation means and generates the analysis image by the analysis image generation means. Done. Therefore, for example, analysis of a layer cross section can be performed with high accuracy.

In order to solve the above-described problem, the layer cross-sectional measurement method of the present invention has at least two sheets arranged at a predetermined interval on the quantity axis correlated with the distance on the scanning line set in the direction intersecting the surface of the laminate . A difference calculation step for calculating a difference between images in units of pixels, and an average value of the calculated differences and a maximum value of the calculated differences for a plurality of pixels forming an array intersecting at least one direction on the image. And a difference statistic value calculation step of calculating a difference statistic value corresponding to each point on the coordinate axis along the one direction, which is at least one of the number of pixels in which the calculated difference is equal to or greater than a threshold value; And an analysis image generation step of generating an analysis image by placing the calculated difference statistical value on a surface stretched between the coordinate axis and the quantity axis .

According to the layer cross-sectional measurement method of the present invention, the change state of the image can be analyzed with higher accuracy as in the case of the layer cross-section measurement system of the present invention described above.

In addition, also in the layer cross-section measuring method of this invention, it is possible to take the various aspects similar to the various aspects in the layer cross-section measuring system of this invention mentioned above.

In order to solve the above problems, the computer program of the present invention causes a computer to function as the above-described layer cross-section measurement system (including various aspects thereof).

According to the computer program of this embodiment, if the computer program is read from a recording medium such as a CD-ROM or DVD-ROM storing the computer program, and executed by a computer provided in the layer cross-section measurement system. Alternatively, if the computer program is executed after being downloaded via communication means, the above-described layer cross-section measurement system according to the present invention can be constructed relatively easily. As a result, the change state of the image can be analyzed with higher accuracy as in the case of the layer cross-section measurement system according to the present invention described above.

  The operation and other advantages of the present invention will become apparent from the best mode for carrying out the invention described below.

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

PBN measurement system and method of the present invention, or the embodiment according to the computer program capable of causing a computer to function as a layer cross-section measurement system will be described with reference to FIGS.

First, with reference to FIG. 1 , the structure of the layer cross-section measuring system which concerns on this embodiment is demonstrated. Figure 1 is a block diagram of a layer cross-section measurement system according to the present embodiment.

In FIG. 1 , the layer cross-section measurement system 200 includes a control unit 200 c, an optical microscope device 201, and a drive device 204. The control unit 200c includes an image memory 202 into which an image from the optical microscope apparatus 201 is captured, and a scanning control device 203 that controls scanning in the driving device 204. In addition , a difference calculating unit 111, a difference statistical value calculating unit 112, An analysis image generation unit 113 and a pattern analysis unit 114 are provided. The layer cross-sectional measurement system 100 further includes a display device 116, a printer 117, and a storage device 118.

  The optical microscope apparatus 201 is an example of the “imaging unit” according to the present invention, and irradiates the multilayer structure 800 with light source light, and optically utilizes the interference between the reflected light and the reference light to optically multilayer the structure. Each image is acquired by receiving minute reflected light from each reflection interface existing in 800. More specifically, the surface image at each depth is perpendicular to the surface of the multilayer structure 800, that is, the focal position or the imaging position is scanned along the scanning direction, with the depth direction of the multilayer structure 800 as the scanning direction. Or it is comprised so that an optical interface image can be imaged.

The image memory 202 stores the plurality of images acquired in this way in association with the scanning lines at least one by one or a plurality of images, that is, sequentially for each depth. More specifically, the image memory 202 stores a plurality of images including at least one image or a plurality of images including two images arranged at a predetermined interval on the quantity axis correlated with the distance. The difference calculation unit 111 is an example of the “difference calculation unit” according to the present invention. The difference calculation unit 111 is stored in the image memory 202 and includes two images arranged at a predetermined interval on the quantity axis correlated with the distance. The difference is calculated in units of pixels. The difference statistical value calculation unit 112 is an example of the “difference statistical value calculation unit” according to the present invention, and the difference calculated by the difference calculation unit 111 for a plurality of pixels forming an array intersecting at least in one direction on the image. Are calculated in association with each point on the coordinate axis along one direction.

  The driving device 204 performs driving to change the focal position or the imaging position of the optical microscope device 201 so as to perform continuous or intermittent scanning at a constant speed along such a scanning direction, and the scanning control device 203 The driving operation by the driving device 204 is configured to be synchronously controlled with the imaging operation in the optical microscope device 201.

  The driving device 204 may be configured to be able to move the optical microscope device 201 along the surface of the multilayer structure 800 (that is, in the X direction or the Y direction). A cross-sectional view of the structure 800 at a desired plane position can be generated as an analysis image finally obtained. However, if the multilayer structure 800 is sufficiently smaller than the plane range that can be imaged by the optical microscope apparatus 201 in the plane size, such as several mm square or several hundred μm square, for example, the optical microscope apparatus 201 is configured in this way. There is no need to move along the surface of the body 800.

For this reason, the image acquisition by the optical microscope apparatus 201 is performed while the optical microscope apparatus 201 is continuously or intermittently or intermittently scanned in a certain direction (that is, the depth direction) by the driving apparatus 204 under the control of the scanning control apparatus 203. Are sequentially performed, and a plurality of images arranged at predetermined or constant intervals on the quantity axis correlated with the distance are acquired. That is, an optical interface diagram at each depth is acquired. Then, thus obtained image or image data, in association with the scanning line in the image memory 202, i.e., Ru stored sequentially in association with each depth.

Here, the calculation of the difference and differential statistical value in the present embodiment, adding explained with reference to FIG. FIG. 2 is a conceptual diagram schematically showing the arrangement of one direction or coordinate axes and pixels in a plurality of cross-sectional images arranged at predetermined intervals on the quantity axis correlated with the distance .

In FIG. 2 , the objective lens 600 of the optical microscope apparatus 201 irradiates light source light such as a thick arrow AR, receives the reflected light, and acquires an optical interface diagram SV using interference with the reference light. At this time, by being driven by the driving device 204, the focal position or imaging position of the objective lens 600 is sequentially changed in the scanning direction (that is, the depth direction or the Z-axis direction) as indicated by the thin arrow SCN in the figure. The optical interface diagrams SV1, SV2,..., SVi,. Here, paying attention to the optical interface diagram SV1, the coordinate axis 601 is defined on the optical interface diagram SV1 while the direction along the x-axis is an example of “one direction” according to the present invention, and the coordinate axis 603 orthogonal thereto is optically defined. Define on interface diagram SV1. Here, in particular, the coordinate axis 601 is defined at the position of the center line in the y direction in the optical interface diagram SV1. In addition, an example of the “quantity axis correlated with the distance ” according to the present invention is defined as the z-axis here. Then, the difference for each pixel Q (x, y, z1) on this optical interface diagram SV is the value of this pixel Q (x, y, z1) (for example, the density for each color) and every certain time. This is a difference from the value of the pixel (x, y, z2) (z2 = z1 + Δz) obtained by shifting by ΔZ in the arrow SCN direction. That is, such a difference is sequentially calculated for each pixel P on the coordinate axis 603 on the optical interface diagram SV1 by the difference calculation unit 111 (see FIG. 1 ), and further, by the difference statistic value calculation unit 112, The difference statistical value for the pixel Q (x, yc, z1) on the coordinate axis 603 and on the coordinate axis 601 is calculated as a difference statistical value using the differences of a plurality of pixels P on the coordinate axis 603. Further, by calculating the difference statistic value of the pixel Q (x, yc, z1) on the coordinate axis 601 with respect to each x, the center line in the y direction in one optical interface diagram SV1 is obtained. A difference statistical value for each pixel Q (x, yc, z1) on the coordinate axis 601 is obtained. Further, by performing the same calculation for a plurality of optical interface diagrams SV1, SV2,..., SVi,..., SVm, the coordinate axis that is the center line in the y direction in each optical interface diagram SVi is finally obtained. A difference statistic value for each pixel Q (x, yc, z1) on 601 is obtained.

Returning to FIG. 1 again, the difference statistical value calculation unit 112 calculates the average value or the maximum value of the differences of the plurality of pixels Q on the coordinate axis 603 as the difference statistical value for the pixel Q (x, yc, z1). It is configured. Alternatively, the number of pixels having a difference equal to or greater than a threshold value among a plurality of pixels Q on the coordinate axis 603 may be calculated as a difference statistical value. The analysis image generation unit 113 is an example of the “analysis image generation unit” according to the present invention, and the difference statistical value calculated by the difference statistical value calculation unit 112 is stretched between the coordinate axis 601 and the quantity axis z. It is configured to generate an image for analysis by placing it on the surface. More specifically, the difference statistical value of each pixel Q (x, yc, z1) on the coordinate axis 601 is stretched between the coordinate axis 601 (in other words, the “direction axis”) and the quantity axis that is correlated with the distance. The numerical value of the pixel of the analysis image to be obtained. Furthermore the analysis image generated by the analysis image generating unit 113 in this embodiment is added with reference to FIG. FIG. 3 is a conceptual diagram schematically showing the analysis image.

In FIG. 3 , the analysis image STM has one y value fixed at the coordinate axis 601 that is the center line in the y direction shown in FIG. 2 , and the x value and the z value are used as parameters. 3 is an analysis image STM of the multilayer structure 800 in the y direction. Here, the difference for a plurality of pixels arranged in the y direction in FIG. 2 is exclusively used to calculate the difference statistical value for the pixel Q (x, yc, z) on the coordinate axis 601. As a result, the signal component in the pixel Q (x, yc, z) is relatively strengthened. In FIG. 3 , the change state pattern PAT is shown as a rough dot shape. However, the actual change state pattern PAT is so thin that the dots cannot be identified with respect to the entire analysis image STM, or the dots It may be a continuous or analog pattern instead of a shape. Returning to FIG. 1 again, the pattern analysis unit 114 is an example of the “pattern analysis unit” according to the present invention. The pattern analysis unit 114 performs pattern analysis on the analysis image DTM generated by the analysis image generation unit 113, and displays the analysis result. It is configured to generate electronic data in a predetermined format. The display device 116 includes, for example, an LCD (liquid crystal display device), a CRT (cathode ray tube), and the like, and displays electronic data as a pattern analysis result by the pattern analysis unit 114 in a predetermined format. The printer 117 is composed of, for example, a laser printer, an ink jet printer, or the like, and prints electronic data as a pattern analysis result by the pattern analysis unit 114 in a predetermined format. The storage device 118 includes, for example, a hard disk device, an optical disk device, and the like, and is configured to store electronic data as a pattern analysis result by the pattern analysis unit 114 in a predetermined format. The control unit 200c may include a difference calculation unit 111, a difference statistical value calculation unit 112, an analysis image generation unit 113, and a pattern analysis unit 114 that have the functions as described above in hardware. Alternatively, installation of the computer program from a recording medium such as a CD-ROM or DVD-ROM storing a computer program that preferably causes a computer having a CPU, memory, etc. to function as each of these units, or download via the Internet, etc. These units may be logically constructed in the computer by installation according to the above.

As shown in FIGS. 2 and 3 , the pixel Q on which the difference statistical value is calculated is limited to only the pixel Q on the coordinate axis 601 fixed on the center line in the y direction (as a result, the coordinate axis 601 Only one analysis image STM corresponding to is finally obtained). That is, the pixel Q on which the difference statistic value is calculated may be the pixel Q on the plurality of coordinate axes 601 arranged at predetermined intervals along the y-axis (as a result, a plurality of pixels corresponding to the plurality of coordinate axes 601 may be obtained. An analysis image STM (y) is finally obtained).

For example, as shown in FIG. 4 , after calculating the difference in the same manner as in FIG. 2 , each pixel Q on the coordinate axis 603 on the optical interface diagram SV1 has a width 604 (but optical You may calculate each using the difference of the some pixel Q in the width | variety shorter than the length of the y direction of an interface diagram. That is, the difference statistic value for a plurality of coordinates Q arranged on the coordinate axis 603 at a predetermined interval or in units of pixels may be calculated using the difference between a plurality of other pixels located within the width 604 around itself. Good. Such operation is performed for a plurality of coordinate axes 603 that are arranged along a coordinate axis 601, further, if this is done for a plurality of optical interface view SV, for example, as shown in FIG. 5, the change state pattern PAT A plurality of analysis images STM (y) each possessed as an image can be generated. In this way, unlike the cases of FIGS. 2 and 3, in the case of FIGS. 4 and 5, a plurality of analysis images STM (y) at a desired y coordinate are obtained not only in the vicinity of the center line. Incidentally, FIGS. 4 and 5, Ru conceptual diagram der in FIGS. 2 and 3 and respectively the same concept.

Here, with reference to FIG. 6 , the analysis image generated by the analysis image generation unit 113 will be described with a specific example. 6 (a) is a plan view showing an analysis image STM generated by the present embodiment, FIG. 6 (b) is a plan view of the control apparatus in the comparative example.

As shown in FIG. 6 (a), an interlayer gap between in the analysis image STM (y) is the layer 701 to 706, each arrangement or, in particular the layer 703 and the layer 705 is a layer 704 is thinner For GP, the layer thickness can be measured with high accuracy. This is considered to be achieved mainly by weakening a highly random noise component in the image data when the difference statistic value calculation unit 112 calculates the difference statistic value.

As shown in FIG. 6 (b), the process of calculating the differential statistical values from the configuration of the present embodiment is removed, according to the comparative example that generated analysis image uses exactly the difference, the layer 701 to 706 The arrangement of each layer is not clear, and particularly in the region where the layer 704 is thin, the ambiguity is remarkable, and the interlayer gap GP cannot be measured with high accuracy. This is considered to be because noise components with high randomness in the image data cannot be weakened mainly when the difference statistic value calculation unit 112 calculates the difference statistic value.

As described above, according to the present embodiment, when the difference statistic value is calculated by the difference statistic value calculation unit 112, a noise component having high randomness in the image data can be weakened. The change state of the image can be analyzed with high accuracy. In addition, the present embodiment is applied to a transparent laminated film in which minute voids and minute foreign matter are assumed to be distributed. At this time, as the difference statistical value, the above-mentioned “maximum value” or “number of pixels equal to or greater than the threshold value” is used. The distribution state in the depth direction can be analyzed.

< Example >
Next, examples of applying the embodiment configured as described above will be described.

The embodiments described above, the embodiment applied to the layer cross-section measurement system will be described with reference to FIGS. FIG. 7 is a block diagram of the system of the present embodiment, and FIG. 8 is an xz layer section and a yz layer section, which are analysis images generated by the system of the present embodiment.

In FIG. 7 , the layer cross-section measurement system 40 is intended to evaluate the transparent multilayer structure sample 43, and the optical microscope apparatus 41 and the focal position or imaging position of the optical microscope apparatus 41 are set to the transparent multilayer structure sample 43. A driving device 44 that sequentially moves in a scanning direction perpendicular to the surface and a computer 45 for measurement and control are provided.

Since it is configured in this way, the imaging position of the optical microscope device 41 is sequentially scanned in the normal direction or depth direction of the surface of the transparent multilayer structure sample 43, or as necessary. Further, by sequentially scanning along the surface, a continuous microscope image of the transparent multilayer structure sample 43 can be observed by the optical microscope device 41. Since the continuous images captured here are acquired by changing the Z axis in time series, the optical interface images at each depth of the transparent multilayer structure sample 43 are represented. At this time, the computer 45 functioning as the difference calculation unit, the difference statistical value calculation unit, the analysis image generation unit, and the scanning control unit described above calculates the absolute value of the difference between the captured images, and further calculates the difference statistical value. by calculating, such illustrated in FIG. 8 (a), and the horizontal axis is the X direction coordinate of the original image and the vertical axis represents the Z direction coordinate of the original image, and the analysis image, FIG. 8 (b 2), an analysis image is created in which the horizontal axis is the coordinate in the Y direction of the original image and the vertical axis is the coordinate in the Z direction of the original image. In these analysis drawings, the portion where the lamination state does not change, that is, the portion where the reflective interface does not exist is basically a flat portion or region where no pattern is drawn, and there is a change in the lamination state. A laminated pattern appears at a place, that is, a place where a reflective interface exists.

  According to the present embodiment, it is possible to measure the position and thickness of a very thin layer such as a micron order, the position and thickness of an interlayer gap or void, etc. with high accuracy by using the analysis image generated using the difference statistics. .

  Note that the present invention is not limited to the above-described embodiment, and can be appropriately changed without departing from the spirit or idea of the invention that can be read from the claims and the entire specification. A state digitizing system and method, and a computer program are also included in the technical scope of the present invention.

It is a block diagram of a layer section measuring system concerning an embodiment of the present invention. It is a conceptual diagram which shows typically an example of the arrangement | positioning of a coordinate axis and a pixel in the some cross-sectional image arranged on a quantity axis of embodiment of this invention . It is a conceptual diagram which shows typically an example of the image for an analysis in embodiment of this invention . It is a conceptual diagram which shows typically the other example of the arrangement | positioning of a coordinate axis and a pixel in the some cross-sectional image arranged on a quantity axis of embodiment of this invention . It is a conceptual diagram which shows the other example of the image for analysis in embodiment of this invention typically. FIG. 6A is a plan view showing an analysis image generated according to the present embodiment , and FIG. 6B is a plan view having the same concept in a comparative example . It is a block diagram of the system of the Example of this invention . It is the xz layer cross section and yz layer cross section which are the images for analysis produced | generated by the system of the Example of this invention.

Explanation of symbols

4 0 ... PBN cut measurement system, 2 00 ... PBN cut measurement system, 2 01 ... optical microscope apparatus, 2 00C ... control unit, 2 02 ... image memory, 203 ... scan controller, 204 ... driving apparatus, 111 ... differential operation , 112... Statistical difference calculation unit, 113... Image generation unit for analysis, 114... Pattern analysis unit , 800.

Claims (5)

A difference calculation means for calculating a difference between at least two images arranged at a predetermined interval on a quantity axis correlated with a distance on a scanning line set in a direction intersecting the surface of the laminate, in units of pixels;
The average value of the calculated difference, the maximum value of the calculated difference, and the number of pixels for which the calculated difference is equal to or greater than a threshold value for a plurality of pixels forming an array intersecting in at least one direction on the image. Differential statistical value calculation means for calculating a differential statistical value corresponding to each point on the coordinate axis along the one direction, which is at least one of them,
A layer cross-section measurement system comprising: an analysis image generation unit configured to generate an analysis image by placing the calculated difference statistical value on a surface stretched between the coordinate axis and the quantity axis.   The layer cross-sectional measurement system according to claim 1, further comprising pattern analysis means for performing pattern analysis on the generated analysis image. It further comprises imaging means for generating a plurality of images by receiving electromagnetic waves including visible light from the laminated body at a plurality of measurement points arranged on the scanning line,
The layer cross-sectional measurement system according to claim 1, wherein the difference calculation unit calculates a difference between the at least two images included in the plurality of images. A difference calculation step of calculating a difference between at least two images arranged at predetermined intervals on a quantity axis correlated with a distance on a scanning line set in a direction intersecting the surface of the laminate, in units of pixels;
The average value of the calculated difference, the maximum value of the calculated difference, and the number of pixels for which the calculated difference is equal to or greater than a threshold value for a plurality of pixels forming an array intersecting in at least one direction on the image. A difference statistic value calculating step of calculating a difference statistic value corresponding to each point on the coordinate axis along the one direction, which is at least one of them,
An analysis image generating step of generating an analysis image by placing the calculated difference statistical value on a surface stretched between the coordinate axis and the quantity axis. A computer program for causing a computer to function as the layer cross-section measurement system according to claim 1 or 2 .

Images