JP5481883B2 - Sample analyzer using specific site detection method - Google Patents

Description

The present invention, use relates sample analyzer using the method of detecting a specific site on a sample, more particularly, how to detect a specific site on the surface of the sample based on the measurement and observation results of the surface shape of the sample The present invention relates to a sample analyzer such as a conventional microscope.

  With the rapid development of technology in the semiconductor field and the micromachine field, the importance of evaluation and analysis of microscopic deformation behavior of metallic and non-metallic materials is increasing. Conventionally, several methods for analyzing distortions and deformations of structures such as bridges and metal samples that are somewhat large are known, and one of them is Digital Image Correlation (DIC). ) (Refer to Non-Patent Documents 1 to 3, Patent Document 1, etc.). In the digital image correlation method, a digital image of a sample before deformation and a digital image of a sample after deformation are acquired, and a correlation coefficient is calculated using luminance (that is, shading) information in units of pixels, so that the image before deformation is calculated. It is grasped to which position on the sample a certain measurement point on the sample has moved after deformation (that is, the same point search before and after the deformation is performed), and information on the deformation is obtained based on this.

  The same point search technique by the digital image correlation method as described above can be applied to an observation image obtained by a microscope. However, in order to apply the digital image correlation method, a certain pattern that can be discriminated by a luminance image is necessary on the surface of the sample. Therefore, it is good if such a pattern is originally on the surface of the sample, but if there is no pattern or if it is unclear, coloring by applying or spraying paint or using laser light, removing partial materials, It is necessary to mark the surface of the sample by a technique such as mechanical deformation such as indentation introduction, that is, to intentionally create a pattern. Therefore, there is a restriction that it is difficult or impossible to perform such marking and the same point search cannot be performed by a digital image correlation method for a sample that does not have a pattern that can be distinguished by a luminance image on the surface.

  On the other hand, in order to observe and measure the surface shape of a small sample placed or accommodated on a sample stage at a very high magnification, a scanning probe microscope (SPM = Scanning Probe Microscope) Various surface analysis apparatuses such as a confocal laser microscope and a digital holographic microscope are known (see Non-Patent Documents 4 and 5). Although these devices have different measurement principles, any of them can acquire the profile of the unevenness of the sample surface, that is, the three-dimensional shape information of the sample surface. However, although it is possible to observe a three-dimensional shape having a high resolution particularly in the vertical direction (height direction) on the sample surface, it is difficult to grasp the deformation state of the sample from this observation result. Moreover, the conventional various apparatuses as described above do not have such a function.

JP 2006-329628 A

TC Chou and three others, “Applications of Digital-Image-Correlation Techniques to Experimental Mechanics”, Experimental・ Mechanics (Experimental Mechanics), Vol. 25, No. 3, pp. 232-244 (1985) Nishikawa, “Full-field microscopic distribution measurement by digital image correlation method”, Non-destructive inspection, Vol. 54, No. 3, pp. 132-138 (2005) B. Pan and three others, “Performance of Sub-Pixel Registration Algorithms in Digital Image Correlation”, Measurement・ Science and Technology, Vol. 17, No. 6, pp. 1615-1621 (2006) Miyamoto, Ito, “Technology and Application of Scanning Laser Microscope / Scanning Probe Microscope Combined Device”, Monthly Tribology, July 2005, pp. 30-33 “DHM 1000 Family”, Aichi Sangyo Co., Ltd., [online], [Search February 25, 2009], Internet <URL: http://www.aichi-sangyo.co.jp/product/lyncee_tec/Lyncee_DHM1000_E. AS.pdf>

The present invention has been made in view of the above problems, and the main object of the present invention is to detect a specific part capable of performing the same point search with high accuracy without intentionally marking the sample surface. The method can be used to evaluate and analyze microscopic deformation behavior on the sample, for example, or unintentional visual field shift due to drift or the like when continuously observing a certain area on the sample It is an object of the present invention to provide a sample analyzer with a new function such as preventing the occurrence of the above.

The specific part detection method used in the sample analyzer according to the present invention made to solve the above problems is a specific part detection method for detecting a specific part on the sample based on the measurement result of the surface shape of the sample. There,
a) a height distribution measuring step for measuring a comparison target height distribution which is a two-dimensional distribution of the surface height of the target sample;
b) a reference height distribution acquisition step for acquiring a reference height distribution which is a two-dimensional distribution of the surface height in a predetermined region, which is a reference for detecting a specific part;
c) The index value indicating the correlation between the reference height distribution and the height distribution in the comparison area having the same size as the predetermined area in the comparison target height distribution is compared with the position of the predetermined area. A height correlation acquisition step for obtaining a two-dimensional distribution of index values indicating a correlation between both height distributions by calculating each of the target height distributions while moving in a two-dimensional direction on the surface of the region,
d) using a two-dimensional distribution of index values indicating the correlation between the two height distributions, finding a position that gives the maximum correlation in the comparison target height distribution and recognizing the position as a specific part Steps,
It is characterized by having.

  Here, the specific site is, for example, a site of interest specified by the user himself / herself (measurer or the like), convenient for characterizing the sample, or for examining behavior such as deformation of the sample. For example, a part that is automatically set for convenience.

  Similar to the height distribution measurement step, the above reference height distribution acquisition step measures the two-dimensional distribution of the surface height of the target sample (that is, the three-dimensional distribution of the sample surface shape) to obtain the height distribution. Alternatively, the reference height distribution may be obtained by reading the data from a storage unit or the like in which the reference height distribution data is created and stored in advance.

In the specific part detection method used in the present invention, a conventionally known digital image correlation method can be used to calculate an index value indicating the correlation from two height distributions. That is, in the specific part detection method used in the present invention, the height correlation acquisition step includes a digital image correlation method for height distribution image data composed of height values of each minute position on the sample surface. By applying an equivalent algorithm, it is possible to obtain a distribution of correlation coefficients as index values indicating correlation.

  As described above, in the digital image correlation method, the deformation or distortion of an object is examined by obtaining the correlation between two digital images in which pixel values (luminance values) are two-dimensionally arranged. In such a specific part detection method, the distribution of the correlation coefficient may be obtained by applying an algorithm equivalent to the digital image correlation method to an image in which the pixel value is replaced with a height value, that is, a height distribution image, as a processing target.

However, if the entire sample surface of the target sample is tilted or curved when the height distribution measurement is performed in the height distribution measurement step, the obtained height distribution indicates the original uneven state of the sample surface. This is a factor that impairs accuracy when examining the correlation of local irregularities. Therefore, in the specific part detection method used in the present invention, the plane correction for correcting the inclination and the curvature of the reference plane of the height with respect to the height distribution obtained by the height distribution measurement step and the reference height distribution acquisition step. It is preferable to further include a processing step.

  According to this, since the reference planes of the reference height distribution and the comparison target height distribution match (become the same plane), the height reflecting local unevenness of the original sample surface in the height correlation acquisition step. Distribution correlation can be obtained. As a result, the accuracy of the correlation index value distribution is improved, and the accuracy of searching for a specific part is also improved. For example, when measuring the three-dimensional shape of the sample surface using various microscopes as will be described later, three-dimensional shape data in which the inclination of the entire sample surface has already been corrected may be obtained.

Further, in the specific part detecting method used in the present invention, preferably, a continuous processing step of using the correlation coefficient in the spatially adjacent micro areas to make the correlation coefficient between the adjacent micro areas continuous. It should be included. “Continuation” here can be rephrased as smoothing or smoothing.

  That is, the spatial resolution (horizontal resolution) of the correlation coefficient distribution before the “continuation” is performed is determined by the horizontal spatial resolution of the original height distribution. On the other hand, if continuous processing is performed, the correlation coefficient is obtained not only for the coordinate position where the height information is obtained in the height distribution but also for the position where the height information is not actually obtained. Therefore, an effect of substantially increasing the spatial resolution in the horizontal direction can be obtained. As a result, it is possible to find the position that gives the maximum correlation coefficient with high accuracy, and the accuracy of the same point search is also improved.

By using the specific site detection method described above , various analyzes on the sample are possible. The analysis here includes not only measurement / measurement of various physical quantities related to the sample but also observation of the sample surface, identification of the sample (determination of whether or not a certain sample is the same as another sample), and the like.

The sample analyzer of the first aspect of the present invention using the above-described specific site detection method is:
A height distribution measuring means for measuring a two-dimensional distribution of the surface height of the sample;
In the two-dimensional distribution of the surface height obtained by measuring the target sample with the height distribution measuring means at the first time point, the specifying means for the user to specify the point of interest ;
Reference height distribution storage means for acquiring and storing the reference height distribution by setting an attention area around the attention point designated by the designation means as the predetermined area;
For the comparison target height distribution obtained by measuring the target sample by the height distribution measuring means at the second time point when the time has elapsed from the first time point, the reference height is obtained by the processing by the height correlation obtaining step. A height correlation calculation means for obtaining a distribution of an index indicating a correlation with the reference height distribution stored in the height distribution storage means;
The same point detecting means for finding the position of the target point on the target sample at the second time point using the index distribution indicating the correlation obtained by the height correlation calculating means,
It is characterized by having.

  In the sample analyzer of the first aspect and each of the following aspects, the height distribution measuring means can be, for example, a digital holographic microscope, a scanning confocal laser microscope, or a scanning probe microscope. As a matter of course, it is desirable that the spatial resolution in the vertical direction, that is, the height direction of the sample surface is high. For example, if the spatial resolution in the height direction is nanometer level, the atomic distribution is reflected in the height distribution (uneven profile on the sample surface), so sample identification that reflects not only the shape but also the composition of the sample is possible. is there.

  In the sample analyzer of the first aspect, the same point search is executed for the same sample (target sample) at the first time point and the second time point after that. For example, when the measurer designates a certain position as the attention point in the height distribution of the sample surface at the first time point, the attention point is searched on the same sample at the second time point after that. For example, after observing one sample with the sample analyzer at the first time point, the sample is temporarily removed, and then when the same sample is mounted on the sample analyzer and observed again, the mounting position of the sample is shifted. Even in this case, it is possible to promptly find the observation point at the first time point and perform observation.

For this purpose, the sample analyzer of the first aspect is
An image acquisition means for acquiring a surface observation image of the sample;
Specific position display means for displaying the position of the point of interest detected by the same point detection means on the surface observation image of the target sample acquired by the image acquisition means at a second time point;
It is good to set it as the structure further provided.

Moreover, the sample analyzer of the first aspect includes:
An image acquisition means for acquiring a surface observation image of the sample;
Moving means for moving at least one of the sample and the image acquisition means in order to change the position of the surface observation image by the image acquisition means;
When a surface observation image of the target sample is obtained by the image acquisition unit at a second time point, the attention point detected by the same point detection unit is located at a predetermined position on the image or within a predetermined range. Tracking control means for controlling the moving means;
It can also be set as the structure further provided.

  The moving means is generally means for moving a sample stage on which a sample is placed or holding a sample. According to the sample analyzer configured as described above, for example, the position tracking is automatically performed so that even if a location on the sample of interest of the measurer moves, the location can be observed. Therefore, for example, as described above, even when the sample is once removed from the apparatus and mounted again, the observation field is set so that the part that has been observed before removal can be automatically searched and observed. be able to. In addition, even if the position of the point of interest on the sample moves over time, such as deformation, expansion / contraction of the sample, or spontaneous movement of the observation object, always keep it from observing the observation field. It is also possible to follow.

In addition, the sample analyzer according to the second aspect of the present invention using the above-described specific site detection method,
A height distribution measuring means for measuring a two-dimensional distribution of the surface height of the sample;
Designation means for the user to designate a plurality of points of interest in the two-dimensional distribution of surface height obtained by measuring the target sample with the height distribution measuring means at the first time point ;
Reference height distribution storage means for acquiring and storing the reference height distribution by setting an attention area around a plurality of attention points designated by the designation means as the predetermined area;
For the comparison target height distribution obtained by measuring the target sample by the height distribution measuring means at the second time point when the time has elapsed from the first time point, the reference height is obtained by the processing by the height correlation obtaining step. Height correlation calculating means for respectively obtaining distributions of indices indicating correlations with a plurality of reference height distributions stored in the height distribution storage means;
The same point detecting means for finding the positions of a plurality of points of interest on the target sample at the second time point using the index distribution indicating the correlation,
A movement amount and a movement direction of the same attention point are calculated from the positions of the plurality of attention points at the first time point and the positions of the plurality of attention points found by the same point detecting means, Displacement detection means for obtaining information on deformation and displacement of the sample from the time point to the second time point;
It is characterized by having.

According to the sample analyzer of the second aspect, when a deformation is caused by applying a mechanical load to the sample, or when the sample expands or contracts due to a stress other than a mechanical load such as heat. The analysis and evaluation of the deformation of the sample can be performed quantitatively.

According to the specific part detecting method used in the present invention, it is possible to search for the same point or the same region by using the information on the surface height that the sample originally has. In general, the surface of a sample made of various materials, whether metal or non-metal, has microscopically clear irregularities even if it has been processed or processed for smoothing. To do. Therefore, unlike the same point search using the conventional digital image correlation method, the same point can be detected even on the sample surface that cannot be identified by the luminance image. Therefore, it is not necessary to mark the sample with coloring, material removal, mechanical deformation, etc., and labor and time for such marking can be saved. Further, the same point search with high accuracy can be performed for a sample that is difficult or impossible to be marked.

Therefore , according to the sample analysis apparatus using the specific part detection method according to the present invention, a new function which has not been provided in the past, in various surface analysis apparatuses such as a scanning probe microscope, a scanning confocal laser microscope, and a digital holographic microscope. Can be added.

The flowchart which shows the implementation procedure of the same site | part detection process which is one form of the specific site | part detection method used for this invention. The conceptual diagram of a plane correction process. The conceptual diagram of the attention point and attention area in reference | standard height distribution. The conceptual diagram of the movement of a comparison area | region in a scanning area | region, and correlation calculation. The conceptual diagram of the continuous process of a correlation coefficient. Explanatory drawing of the continuous process of a correlation coefficient. The figure which shows the surface uneven | corrugated profile of the different site | part of the same sample. Explanatory drawing of an example of the same area search using a no-load test piece. Explanatory drawing of an example of the same area search using a no-load test piece. Explanatory drawing of an example of the same area search using a load test piece. Explanatory drawing of an example of the same area search using a load test piece. Schematic block diagram of an embodiment of a locking Ru specimen analyzing apparatus according to the present invention. Schematic block diagram of another embodiment of the engagement Ru specimen analyzing apparatus according to the present invention. Explanatory drawing of an example of the measurement operation | movement in the sample analyzer of one Example.

First, an embodiment of the specific part detection method used in the present invention will be described. In this embodiment, the same part detection process for searching for a point or region on the comparison target sample that is the same as a certain point or region on the reference sample is performed.

  Note that the “reference sample” and the “comparison sample” herein may be different samples, but may be the same sample. For example, in measurement of deformation or movement of a sample, the sample before such deformation or movement is a reference sample, and the sample after deformation or movement is the same sample as the sample to be compared. It becomes. Also, in applications such as checking whether the comparison sample is a genuine product or a counterfeit product, the reference sample and the comparison sample are different samples, and the reference sample is a regular product or a standard product equivalent to the regular product. It is.

FIG. 1 is a flowchart showing an execution procedure of the same part detection process according to this embodiment.
First, as a first step, a two-dimensional distribution of the height of the surface of the reference sample and a two-dimensional distribution of the height of the surface of the comparison target sample are acquired (steps S1 and S2). Here, the two-dimensional distribution of the height of the comparison target sample surface is referred to as a comparison target height distribution. In order to acquire such a height distribution on the sample surface, the above-described digital holographic microscope, scanning confocal laser microscope, scanning probe microscope, or the like can be used. The height information of the sample surface is information at a position having discrete coordinates on the x-axis and the y-axis that are orthogonal to each other and that form the xy plane that is a horizontal plane. Accordingly, the height distribution becomes a three-dimensional histogram as shown in FIG. 4A, for example.

  Next, a plane correction process is performed on the two height distributions acquired in steps S1 and S2 in order to align the height reference plane (step S3). This is a process for excluding an element different from the local unevenness of the sample surface in question, for example, the inclination of the entire sample surface or a large curvature of the sample surface different from the local unevenness. . 2A and 2B are diagrams showing the concept of the plane correction process. FIG. 2A is a diagram showing a height distribution before plane correction, and FIG. 2B is a diagram showing a height distribution after plane correction.

Now, as shown in FIG. 2A, it is assumed that the reference plane of the height distribution is inclined with respect to the x′-y ′ plane. At this time, for example, the following arithmetic processing can remove the influence of the average inclination that does not reflect the essence of the height distribution.
That is, the reference plane G _std (x ′, y ′) can be approximated by the following equation (1).
G _std (x ′, y ′) = Lx ′ + my ′ + n (1)
The actual height value at coordinates (x _i ′, y _i ′) on the x′−y ′ plane is G (x _i ′, y _i ′), and the value on the reference plane G _std (x ′, y ′) is _Assuming G _std (x _i ′, y _i ′), the height d _i at each coordinate after plane correction is
d _i = G (x _i ′, y _i ′) −G _std (x _i ′, y _i ′)
It is. Since the influence of the plane inclination is removed in a state where the plane correction is performed so that the reference plane G _std (x ′, y ′) coincides with the x′−y ′ plane, the difference d between the reference plane and the actual height value d. _It is reasonable to think that the sum of squares of _i is minimized. Therefore, the constants L, m, and n in equation (1) are determined so that the square sum of d _i is minimized. That is, L, m, and n are obtained by solving simultaneous linear equations related to constants L, m, and n expressed by the following equation (2).
(∂ / ∂L) Σd _i ² = (∂ / ∂m) Σd _i ² = (∂ / ∂n) Σd _i ² = 0 (2)
(Where Σ represents the sum of i)
The least square method is well known and will not be described in detail here.
Since the reference plane G _std (x ′, y ′) is obtained by the above calculation, the height G (x _n ′, y _n ′) of each coordinate position is corrected based on the reference plane G _std (x ′, y ′). A height distribution as shown can be obtained.
In addition, when the reference plane is not simply inclined with respect to the x′-y ′ plane but is curved, the plane correction is complicated, but in many cases, only the above-described inclination is considered. It is enough.

  Next, a point of interest (POI) and a region of interest (ROI) are set in the reference height distribution subjected to the plane correction, as shown in FIG. 3 (step S4). Now, consider a case where the attention area is rectangular and the attention point is set at the approximate center of the attention area. If the region of interest is too small, it is determined that the correlation is high in the correlation coefficient calculation described later, even though it is not the same part as the comparison region, and the possibility of false detection increases. Conversely, if the region of interest is too wide, false detection can be avoided, but an oversight that is judged to have a low correlation even though it is actually the same part is likely to occur. Therefore, it is desirable that the size of the region of interest is appropriately determined according to the uneven state of the sample surface at that time. In the example of FIG. 3, when the attention area shown in the figure is set, a height distribution indicated by diagonal lines is extracted, and this becomes a reference height distribution.

  Subsequently, as shown in FIG. 4A, a scanning region is set in the comparison target height distribution (step S5). The scanning region is a range in which a region having a height distribution similar to that of the region of interest (that is, having a high correlation) is searched for in the region in which the comparison target height distribution is obtained. As the scanning region is wider, the amount of calculation for calculating the correlation coefficient and the like increases, so that it takes time to search for the same part. On the other hand, since the search for the same part is performed only within the range of this scanning region, for example, when it is desired to find the same point whose position has been moved due to deformation of the sample, the scanning region is surely included in the moving destination. Need to be set. Therefore, when the amount of movement and the direction of movement are unknown, the scanning area is set wider with some margin.

  Then, as shown in FIG. 4A, a comparison region having the same size on the x′-y ′ plane as the region of interest ROI is set in the comparison target height distribution, and this comparison region is set as a scanning region. While moving discretely within the range, that is, step by step in coordinate position units (step S6), the comparison target height distribution included in the comparison area is compared with the reference height distribution included in the attention area, A correlation coefficient of the height distribution is calculated (step S7). In the example of FIGS. 3 and 4, the number of reference height distribution data included in the region of interest ROI is nine in total, three each in the x-axis direction and the y-axis direction (see FIG. 4B). . Accordingly, the height data included in the comparison area is similarly nine, and as shown in FIG. 4A, nine comparison target height distribution data adjacent to the comparison area in the x-axis direction and the y-axis direction. It is moved sequentially step by step so that it is included.

The same algorithm as the digital image correlation method is used for calculating the correlation coefficient between the reference height distribution data and the same number of comparison target height distribution data, and for searching for the same point based on the calculation result. Take an example. Now, assuming that the function representing the reference height distribution is F and the function representing the comparison target height distribution is G, a correlation coefficient R (u, v) representing the strength of the correlation between the function F and the function G is, for example, It calculates | requires by following (3) Formula. The range of values that the correlation coefficient can take is −1 ≦ R (u, v) ≦ 1.
Here, S is the attention area, (x ′ _L , x ′ _L ) is the coordinates in the attention area S, and F _m and G _m are the average values of the functions F and G, respectively. Further, (x ′ _L + u, y ′ _L + v) is a coordinate after movement of the coordinates (x ′ _L , x ′ _L ).

  As described above, each time the comparison area is moved within the scanning area, the correlation coefficient R (u, v) is obtained by the above equation (3), so that the area is slightly smaller than the scanning area as shown in FIG. A two-dimensional distribution of correlation coefficient values as shown is obtained. This two-dimensional distribution is a value for each coordinate position on the xy plane. If the height distribution, that is, the spatial resolution in the horizontal direction of the three-dimensional measurement of the sample surface is sufficiently high, the two-dimensional distribution of the correlation coefficient is substantially curved. However, in general, the spatial resolution in the horizontal direction of the sample surface three-dimensional measurement by the various surface analyzers as described above is lower than the spatial resolution in the vertical direction. Therefore, using the correlation coefficient obtained as described above, the correlation coefficient at continuous (actually more discrete at every minute interval) between adjacent coordinate positions on the xy plane is calculated. By obtaining by smoothing processing, the correlation coefficient is made continuous (step S8). The continuation of the correlation coefficient can be performed using the technique described in Non-Patent Document 3, for example.

An example of continuous correlation coefficients will be described. Now, assume that a certain coordinate position on the xy plane is (u _L , v _L ) = (0, 0), and adjacent coordinate positions are represented by −1, 0, 1 as shown in FIG. Shall. At this time, a continuous correlation coefficient between a coordinate position where (u _L , v _L ) = (0, 0) and a coordinate position adjacent thereto is obtained by the following equation (4).
R (u _L , v _L ) = a ₀ + a ₁ u _L + a ₂ v _L + a ₃ u _L v _L + a ₄ u _L ² + a ₅ v _L ² (4)
Here, −1 ≦ R (u _L , v _L ) ≦ 1, −1 ≦ u _L ≦ 1, and −1 ≦ v _L ≦ 1. a ₀ , a ₁ , a ₂ , a ₃ , a ₄ , a ₅ are coefficients determined by the method of least squares, and the value obtained from the above-mentioned correlation coefficient distribution is calculated by putting it in the equation (4). can do.

  By performing such correlation coefficient continuation processing, the two-dimensional distribution of correlation coefficients as shown in FIG. 5A is approximated by a smooth curved surface as shown in FIG. Thereby, even if the spatial resolution in the horizontal direction of the three-dimensional measurement of the sample surface is somewhat deteriorated, a correlation coefficient distribution having a high horizontal resolution can be obtained. In the correlation coefficient distribution thus obtained, a position that gives the maximum correlation coefficient is searched (step S9). As described above, since the correlation coefficient distribution is continuous, the position where the correlation coefficient is maximized is not necessarily on discrete coordinate positions. If the position that gives the maximum correlation coefficient is found, it is determined that it is the same point as the point of interest POI in the sample to be compared, and the region surrounding it is the region of interest ROI (step S10).

However, when the maximum value of the correlation coefficient is a small value or when a clear peak that gives the maximum value of the correlation coefficient does not appear, it may be determined that there is no identical point and identical region.
Through the procedure described above, it is possible to accurately find a point or region that is the same as the point of interest and the region of interest set on the surface of the reference sample on the surface of the comparison target sample.

Next, an experimental example in which the same part detection process is applied to an actual sample will be described.
In this experiment, a 5 mm × 5 mm × 10 mm square columnar pure titanium test piece was used as a sample, and mechanical polishing, heat treatment, electrolytic polishing, which are surface treatments usually performed before performing a mechanical test, And chemical corrosion. FIG. 7 shows an example of the surface unevenness profile of different parts of the same sample, measured using a digital holographic microscope (DHM R1000 manufactured by Lynce Tec, Switzerland).

  In the range shown in this figure, one side is about 20 to 30 μm, and the difference in height is about several tens of nm. Even when the surface treatment for planarization as described above is performed, a clear unevenness profile appears in the nanometer order, and the unevenness profile is different for each part. That is, this uneven profile can be handled in the same way as a human fingerprint, and by using this, it is possible to search for the same point on the sample and to verify and authenticate that it is the same sample.

FIG. 8 is an image showing the result of performing the same point search by the above method based on the surface unevenness profile of the same sample in an unloaded state, measured using the digital holographic microscope. A surface observation image having a height distribution, and (b) is a surface observation image having a comparison target height distribution. The same point search conditions are: size of all measurement areas: 700 × 700 pixels, size of attention area ROI: 40 × 40 pixels, attention point POI: center point of attention area, scanning area size: 80 × 80 pixels. . In this example, the same point search is performed for 23 attention points set on the reference height distribution, but the same portion is detected on the comparison target height distribution for all of them. 9A and 9B are three-dimensional measurement images at this time. FIG. 9A is an image around the target point [1] of the reference height distribution, and FIG. 9B is an image around the search result point in the comparison target height distribution. . From this figure, it can be seen that both have the same uneven pattern.
From this result, it can be confirmed that the same part searching process is effective.

On the other hand, FIG. 10 is an image showing the result of performing the same point search by the above method based on the surface unevenness profile of the same sample before and after the compression load test, and (a) is a reference height distribution (before the load test). ), (B) is the comparison target height distribution (after the load test). The same point search condition is the same as in the case of no load described above. In this example as well, the same point search is performed for 23 attention points set on the reference height distribution, but it can be seen that similar portions are accurately detected on the comparison target height distribution. . FIG. 11 is a three-dimensional measurement image at this time, where (a) is an image around the target point [1] of the reference height distribution, and (b) is an image around the search result point in the comparison target height distribution. . From this figure, it can be seen that the pattern is very similar.
From the above, it can be confirmed that the same part searching process is sufficiently effective even in a state in which a load is applied to the sample and a certain degree of deformation occurs.

Next, an example of the sample analyzer according to the present invention using the above-described specific site detection method will be described. FIG. 12 is a schematic block diagram of the sample analyzer according to this embodiment.

  The sample analyzer acquires three-dimensional shape data for the sample 2 placed on the sample stage 1 and also measures the three-dimensional shape data obtained by the measuring unit 3 that captures and images the three-dimensional shape data. A three-dimensional measurement data processing unit 5 that forms a height distribution, an observation image processing unit 4 that processes an image signal captured by the measurement unit 3, a reference height information storage unit 7 that stores a reference height distribution, The height correlation processing unit 6 that calculates the continuous correlation coefficient distribution as described above from the reference height distribution and the comparison target height distribution, and searches for the same point or the same region based on the correlation coefficient distribution A stage driving unit 9 including a same point detection unit 8 and a motor for moving the sample stage 1 within a predetermined range at least in the x-axis direction and the y-axis direction (usually also in the z-axis direction perpendicular to these two axes). A control unit 10 that executes various controls; It includes an operation portion 11 for over THE operates, a display unit 12, a. The measurement unit 3 may be of any measurement method as long as it can acquire the three-dimensional shape data of the surface of the sample 2. For example, measurement by a scanning probe microscope, a scanning confocal laser microscope, or a digital holographic microscope is possible. Techniques can be used.

The characteristic operation of this sample analyzer will be described below with reference to FIG.
That is, when the sample 2 is placed on the sample stage 1, the measurement unit 3 performs measurement on the sample 2, and the three-dimensional measurement data processing unit 5 performs a predetermined range on the sample 2 based on the three-dimensional measurement data. The uneven profile (height distribution) is obtained. In parallel with this, the observation image processing unit 4 forms an optical observation image of a predetermined range on the sample 2, for example. The optical observation image is displayed on the screen of the display unit 12 by the control unit 10, and the measurer designates an observation point by the operation unit 11 while viewing the observation image.

  Now, for example, as shown in FIG. 14A, the sample 2 is placed in the sample placement range 30 on the sample stage 1, and only the range of the observation visual field 31 therein is displayed as an optical observation image on the display unit 12. It is assumed that it is displayed on the screen. While the position of the observation visual field 31 is fixed, the sample placement range 30 and the sample 2 can be moved with the movement of the sample stage 1, so when the sample stage 1 is moved, the observation visual field 31 is moved to the observation visual field 31 on the sample 2. The part that fits changes. That is, the measurer can observe an arbitrary part on the sample 2 by appropriately moving the sample stage 1. Further, the size of the observation visual field 31 differs depending on the magnification, and the observation visual field 31 becomes smaller as the magnification is increased. Here, in FIG. 14A, it is assumed that the position indicated by the point 32 is set as an observation point by the measurer.

  When the observation point is designated, the control unit 10 obtains a coordinate position corresponding to the observation point on the stage coordinates for determining the position of the sample stage 9 and notifies the height correlation processing unit 6 of the coordinate position. The height correlation processing unit 6 obtains a position corresponding to the observation point in the height distribution, and sets an attention area of a predetermined size around the position as an attention point. Then, a height distribution that falls within the region of interest is extracted and stored in the reference height information storage unit 7 as a reference height distribution.

  Once the measurer has finished measuring and observing the sample 2, once the sample 2 is taken out of the sample stage 1, the same sample (if the sample is the same, even if there is some deformation due to a load test or the like) Consider a case where 2) is placed on the sample stage 1 for observation. At this time, in many cases, the mounting position of the sample 2 on the sample stage 1 is not the same as that at the time of the first measurement, and a positional shift (here, xy) as shown in FIG. Assuming a positional shift in a state of translation on the surface). At this time, even if the measurer wishes to observe the previous observation point again, the observation point may be out of the observation field of view 31, and the observation image can be found where the observation point is. Can be difficult.

  Therefore, when the measurer instructs the execution of the observation point tracking using the operation unit 11, the height correlation processing unit 6 compares the height distribution of the surface of the sample 2 created by the three-dimensional measurement data processing unit 5 at that time. A target height distribution is obtained, and a correlation coefficient distribution between this and a reference height distribution stored in the reference height information storage unit 7 is obtained. The same point detection unit 8 searches for the same point with respect to the target point from the correlation coefficient distribution, and informs the control unit 10 of the coordinate position of the same point. Since the coordinate position of the same point is the coordinate position of the current observation point, the tracking movement amount calculation unit 100 in the control unit 10 moves the same point to the center of the observation field 31 in the x-axis direction and the y-axis direction. And the stage drive unit 9 is controlled accordingly to move the sample stage 1. As a result, as shown in FIG. 14C, the sample placement range on the sample stage 1 is 30a → 30b, the sample is moved from 2a → 2b, and the observation point is an abbreviation of the observation field 31 as 32a → 32b. Move to the center. In this way, the observation point on the sample 2 can be kept within the range of the observation visual field 31.

  The above description is an example of the case where the sample placement position has shifted due to the same sample being placed on the sample stage 1, but even if the sample itself or the observation point on the sample moves due to some factor. The observation point can be followed in the same manner as described above. The movement of the observation point on the sample itself or the sample can be both spontaneous movement and movement due to an external factor. The former includes, for example, the movement and operation of a micromachine. Examples of the latter include expansion / compression due to heat and the like. In addition, there is no change in the position of the sample stage 1 or the sample 2 itself, and there may be a case where the observation field of view 31 itself is shifted (drifted) due to a temperature change or the like in the measuring unit 3 and deviates from the observation point. The observation visual field can be made to follow the observation point in the same manner as described above.

  The example shown in FIG. 14 assumes a positional shift in which the sample moves in parallel on the xy plane. However, if the component is small even if there is a positional shift due to rotation around the z axis, This can be handled only by movement in the x-axis direction and the y-axis direction. On the other hand, when the rotational component around the z-axis is large to some extent, the sample stage 1 is configured to be rotatable around the z-axis and rotated by a predetermined minute angle around the z-axis when searching for the same part. What is necessary is just to add the process which calculates | requires correlation coefficient distribution at the time.

FIG. 13 is a schematic block diagram of a sample analyzer related to the present invention although not included in the present invention . Although the configuration is basically the same as that of the sample analyzer shown in FIG. 12, a regular product height information storage unit 20 is provided instead of the reference height information storage unit 7, and the control unit 10 tracks the tracking movement amount calculation unit 100. Instead, a regular product identification unit 101 is provided.

  In this sample analyzer, a characteristic irregularity profile is formed in advance on a regular product of a certain type of product, and the product is genuine by checking whether or not the irregularity profile is the same as this irregularity profile. Or a counterfeit product. In this case, the characteristic uneven profile is stored in the regular product height information storage unit 20 as a reference height distribution. When the product to be inspected is placed on the sample stage 1 as the sample 2 and the execution of the inspection is instructed, the height correlation processing unit 6 causes the surface of the sample 2 created by the three-dimensional measurement data processing unit 5 at that time. And a correlation coefficient distribution between this and the reference height distribution stored in the regular product height information storage unit 20 is obtained. The same point detection unit 8 searches the correlation coefficient distribution for a region that is the same as the characteristic uneven profile, and notifies the control unit 10 of the search result.

  Under the assumption that all regular products have the above-described characteristic uneven profile, if this uneven profile cannot be found in a certain product, it can be determined that this product is highly likely to be a counterfeit product. Therefore, when the search result that the same region as the characteristic uneven profile is found is obtained in the same point detection unit 8, the regular product identification unit 101 identifies that the inspected sample 2 is a regular product. The result is displayed on the display unit 12. On the other hand, when a search result is obtained that an area identical to the characteristic uneven profile cannot be found, the genuine product identification unit 101 displays the identification result that the inspected sample 2 is a counterfeit product. To display.

  Further, the confirmation / evaluation of products using the characteristic uneven profile as described above can be used for various purposes other than the identification of genuine / counterfeit products. For example, by forming a unique uneven profile on the product surface for each type of product, or even for the same product, for each manufacturing plant, this characteristic uneven profile can be identified to identify the product type and manufacturing factory. It becomes possible to specify. In other words, the uneven profile can be used as a product tag or used for product traceability.

  Furthermore, as shown in the above experimental example, a plurality of points of interest are set on the sample surface before the load test, and a point that is the same as each of the points of interest after the load test is searched for. If the amount of movement and the direction of movement are determined, quantitative evaluation of the deformation of the sample by a load test can be performed.

  In addition, since the said Example is an example of this invention, even if it corrects, changes, an addition etc. suitably in the range of the meaning of this invention in points other than what was described above, it is included by this invention. Is clear.

DESCRIPTION OF SYMBOLS 1 ... Sample stage 2 ... Sample 3 ... Measurement part 4 ... Observation image processing part 5 ... Three-dimensional measurement data processing part 6 ... Height correlation processing part 7 ... Reference height information storage part 8 ... Same point detection part 9 ... Stage drive Unit 10 ... Control unit 100 ... Tracking movement amount calculation unit 101 ... Genuine product identification unit 11 ... Operation unit 12 ... Display unit 20 ... Genuine product height information storage unit

Claims (8)

A sample analyzer using a specific part detection method for detecting a specific part on the sample based on the measurement result of the surface shape of the sample ,
The specific site detection method includes:
a) a height distribution measuring step for measuring a comparison target height distribution which is a two-dimensional distribution of the surface height of the target sample;
b) a reference height distribution acquisition step for acquiring a reference height distribution which is a two-dimensional distribution of the surface height in a predetermined region, which is a reference for detecting a specific part;
c) The index value indicating the correlation between the reference height distribution and the height distribution in the comparison area having the same size as the predetermined area in the comparison target height distribution is compared with the position of the predetermined area. A height correlation acquisition step for obtaining a two-dimensional distribution of index values indicating a correlation between both height distributions by calculating each of the target height distributions while moving in a two-dimensional direction on the surface of the region,
d) using a two-dimensional distribution of index values indicating the correlation between the two height distributions, finding a position that gives the maximum correlation in the comparison target height distribution and recognizing the position as a specific part Steps,
I have a,
A height distribution measuring means for measuring a two-dimensional distribution of the surface height of the sample;
Designation means for the user to designate a point of interest in the two-dimensional distribution of surface height obtained by measuring the target sample with the height distribution measurement means at the first time point;
Reference height distribution storage means for acquiring and storing the reference height distribution by setting an attention area around the attention point designated by the designation means as the predetermined area;
For the comparison target height distribution obtained by measuring the target sample by the height distribution measuring means at the second time point when the time has elapsed from the first time point, the reference height is obtained by the processing by the height correlation obtaining step. A height correlation calculation means for obtaining a distribution of an index indicating a correlation with the reference height distribution stored in the height distribution storage means;
The same point detecting means for finding the position of the target point on the target sample at the second time point using the index distribution indicating the correlation obtained by the height correlation calculating means,
Sample analysis apparatus comprising: a. A sample analyzer using a specific part detection method for detecting a specific part on the sample based on the measurement result of the surface shape of the sample ,
The specific site detection method includes:
a) a height distribution measuring step for measuring a comparison target height distribution which is a two-dimensional distribution of the surface height of the target sample;
b) a reference height distribution acquisition step for acquiring a reference height distribution which is a two-dimensional distribution of the surface height in a predetermined region, which is a reference for detecting a specific part;
c) The index value indicating the correlation between the reference height distribution and the height distribution in the comparison area having the same size as the predetermined area in the comparison target height distribution is compared with the position of the predetermined area. A height correlation acquisition step for obtaining a two-dimensional distribution of index values indicating a correlation between both height distributions by calculating each of the target height distributions while moving in a two-dimensional direction on the surface of the region,
d) using a two-dimensional distribution of index values indicating the correlation between the two height distributions, finding a position that gives the maximum correlation in the comparison target height distribution and recognizing the position as a specific part Steps,
I have a,
A height distribution measuring means for measuring a two-dimensional distribution of the surface height of the sample;
Designation means for the user to designate a plurality of points of interest in the two-dimensional distribution of surface height obtained by measuring the target sample with the height distribution measuring means at the first time point;
Reference height distribution storage means for acquiring and storing the reference height distribution by setting an attention area around a plurality of attention points designated by the designation means as the predetermined area;
For the comparison target height distribution obtained by measuring the target sample by the height distribution measuring means at the second time point when the time has elapsed from the first time point, the reference height is obtained by the processing by the height correlation obtaining step. Height correlation calculating means for respectively obtaining distributions of indices indicating correlations with a plurality of reference height distributions stored in the height distribution storage means;
The same point detecting means for finding the positions of a plurality of points of interest on the target sample at the second time point using the index distribution indicating the correlation,
A movement amount and a movement direction of the same attention point are calculated from the positions of the plurality of attention points at the first time point and the positions of the plurality of attention points found by the same point detecting means, Displacement detection means for obtaining information on deformation and displacement of the sample from the time point to the second time point;
Sample analysis apparatus comprising: a. The sample analyzer according to claim 1 or 2,
Said height correlation acquisition step of the specific portion detecting method, with respect to the height distribution image data composed of the value of the height of each micro position on the sample surface, applying a digital image correlation method algorithm equivalent the sample analyzer and obtaining a distribution of correlation coefficient as an index value indicating the correlation. The sample analyzer according to claim 3,
The specific part detecting method includes a plane correction processing step of correcting the inclination and / or bending of the reference surface of the height distribution obtained by the height distribution measurement step and the reference height distribution acquisition step. A sample analyzer further comprising: The sample analyzer according to claim 3,
The specific part detection method further includes a sample processing step of using a correlation coefficient in spatially adjacent microregions to make a correlation coefficient between the adjacent microregions continuous. Equipment . The sample analyzer according to claim 1 ,
An image acquisition means for acquiring a surface observation image of the sample;
Specific position display means for displaying the position of the point of interest detected by the same point detection means on the surface observation image of the target sample acquired by the image acquisition means at a second time point;
A sample analyzer further comprising: The sample analyzer according to claim 1 ,
An image acquisition means for acquiring a surface observation image of the sample;
Moving means for moving at least one of the sample and the image acquisition means in order to change the position of the surface observation image by the image acquisition means;
When a surface observation image of the target sample is obtained by the image acquisition unit at a second time point, the attention point detected by the same point detection unit is located at a predetermined position on the image or within a predetermined range. Tracking control means for controlling the moving means;
A surface observation apparatus further comprising: The sample analyzer according to any one of claims 1 to 7 ,
The sample analyzer according to claim 1, wherein the height distribution measuring means is any one of a digital holographic microscope, a scanning confocal laser microscope, and a scanning probe microscope.