JP6166702B2 - Length measuring device and length measuring method - Google Patents

Description

  The present invention relates to a length measuring device and a length measuring method, and more particularly, to a length measuring device and a length measuring method for estimating a concavo-convex shape of a sample surface and a position change due to aging, thermal expansion / shrinkage, swelling, and the like.

  As a device for measuring the shape and position of a sample, there are a microscope equipped with an image sensor such as a CCD and a laser displacement meter. FIG. 7 shows the configuration of a conventional length measuring device. The length measuring device 700 includes a light source 701, a sensor 702, a drive unit 703, and a scale 704.

  For example, in the case of a microscope equipped with an image sensor, the length measuring device 700 uses illumination light as input light from the light source 701, an image sensor such as a CCD as the sensor 702, and an optical system including a lens as the drive unit 703. Is moved in the Z direction so as to approach or move away from the sample.

  The contrast of the image obtained by the sensor 702 is calculated as an evaluation amount, and the distance between the sample 710 and the sensor 702 is changed by the driving unit 703 so that this value becomes maximum. In the case where an optical element such as an objective lens is provided, the optical distance may be changed by driving them, or the mechanism on which the sensor 702 is mounted may be moved. By reading the distance when the evaluation amount is maximum from the scale, the distance between the surface of the sample 710 and the imaging surface of the sensor 702 can be measured. By repeating this in the in-plane direction of the sample 710, the surface shape of the sample 710 can be measured.

  In the case of a laser displacement meter, a laser that outputs single-wavelength light is used as the light source 701, a photodetector having a pinhole on the front surface as the sensor 702, and a lens as the driving unit 703 (see Non-Patent Document 2). . The amount of light when the light reflected and returned from the surface of the sample 710 is detected by the sensor 702 via the pinhole is defined as an evaluation amount. The drive unit 703 changes the distance between the surface of the sample 710 and the imaging surface of the sensor 702 so that the evaluation amount is maximized. In this case, the evaluation amount is maximized when the lens is driven to focus on the sample surface. By reading the position of the lens at this time, the distance between the surface of the sample 710 and the imaging surface of the sensor 702 can be measured. The shape can be measured by moving the stage on which the sample 710 is mounted in an in-plane direction perpendicular to the laser beam traveling direction.

Hidenao Kume, “The digital camera market continues to shrink and reversal with the introduction of Toranoko technology”, Nikkei Electronics, 8-9, March 3, 2014 issue “Principles of Confocal Optical Systems” [online], Lasertec Corporation, [searched August 20, 2014], Internet <URL: http://www.lasertec.co.jp/products/ principle / principle2.html>

  However, the conventional measurement method has a problem that the measurement accuracy deteriorates when these methods are applied to a rough surface having irregularities.

  For example, when a microscope is used, if the surface is flat, a curve having a peak image contrast value with respect to the amount of movement in the Z direction is obtained as shown in FIG. is there. However, when there are a plurality of projections and depressions in the range reflected by the sensor 702 on the surface of the sample 710, there are a plurality of distances at which the contrast corresponding to each projection and projection is a maximum as shown in FIG. It is difficult to select a maximum value corresponding to one unevenness of the object from the inside. Therefore, it is not possible to measure the distance to the individual irregularities, but only to obtain a measurement accuracy of the degree of irregularities in the surface unit in the range reflected in the sensor 702.

  Similarly, with respect to the laser displacement system, it is dependent on which part of the concavo-convex surface is focused, and an accuracy higher than that of the concavo-convex included in the laser beam diameter cannot be obtained. Further, since the surface is an uneven surface, the input light is scattered and the amount of light returning to the sensor 702 is small, and the detection accuracy of the sensor 702 is lowered.

  Furthermore, it may be focused locally with a laser or the like and scanned in the horizontal direction of the concavo-convex surface, but it is necessary to adjust the drive unit to focus at each point. There is a problem that it takes time to fall.

  In addition, when the measurement takes time, there is a problem in that the accuracy is further deteriorated due to drift of the measurement device or the sample or thermal deformation during the measurement.

  From the above, it has been difficult to obtain accurate values of the height and distance of a sample having an uneven surface in the prior art. Therefore, even if you try to capture the change in shape and position due to changes over time, thermal expansion / contraction, swelling, etc. by surface displacement, if you focus locally on an uneven surface, you can measure the height, but as a whole surface It was difficult to measure with high accuracy how it was displaced.

  The present invention has been made in view of such problems, and its object is to move from the imaging surface of the image sensor to the surface of the sample without moving the image sensor or the optical element for focusing. The object is to provide a length measuring device and a length measuring method capable of estimating a distance.

  In order to solve the above problems, the present invention provides a length measuring device, an imaging device, a drive unit capable of changing a distance between the imaging device and a table, an imaging surface of the imaging device, and the platform A scale for measuring the distance from the surface to be measured of the sample placed on the surface, and at least part of the surface to be measured of the sample obtained by photographing the surface to be measured of the sample placed on the table with the imaging device Two or more pieces of reference data including an in-focus image and an actual distance value measured at the scale when an image in focus on at least a part of the surface to be measured of the sample is captured, and the imaging A storage unit that stores at least one target data including at least an image obtained by photographing the surface of the sample to be measured, and any of the images of the reference data and the images of the target data stored in the storage unit Image processing range Then, while translating and rotating, calculate the mean square of the difference in gradation value in the overlapping area between each image of the reference data and each image of the target data, the smallest among the translated and rotated The root mean square is the degree of coincidence between each image of the reference data and each image of the target data, and an estimated distance value from the image sensor to the surface to be measured of the sample when each image of the target data is captured. And a calculation unit that estimates based on the degree of coincidence of each of the reference data with each image and the corresponding measured distance value of the reference data.

  According to a second aspect of the present invention, in the length measuring device according to the first aspect, the target data includes an actual distance value measured at the scale when each image of the target data is captured, and the calculation is performed. The unit calculates the displacement of the position of the surface to be measured of the sample as a difference between the estimated distance value and the measured distance value of the corresponding target data.

  According to a third aspect of the present invention, in the length measuring apparatus according to the second aspect, the calculation unit calculates the displacement of the position of the surface to be measured of the sample, and the measured distance value of the target data indicates the target. It is calculated by performing regression analysis on the assumption that the sum of the estimated distance value of the data and the displacement of the position of the surface of the sample to be measured is equal.

  According to a fourth aspect of the present invention, in the length measuring device according to any one of the first to third aspects, when the range of the measured distance value of the reference data is z1 to z2, each image of the target data is The distance measured with the scale is taken at a position between z1 and z2.

  According to a fifth aspect of the present invention, in the length measuring device according to any one of the first to fourth aspects, the sample includes a reference plane different from the surface to be measured, and the scale has zero the reference plane. The distance between the imaging surface of the imaging device and the surface to be measured of the sample is calculated as a reference.

  A sixth aspect of the present invention is the length measuring device according to any one of the first to fifth aspects, wherein the calculation unit includes the estimated distance value of the target data and an image having the smallest degree of coincidence. The measured distance value of the reference data is used.

  The invention according to claim 7 is the length measuring device according to any one of claims 1 to 5, wherein the calculation unit calculates the estimated distance value of each of the target data and the measured distance value of each of the reference data. Is calculated by adding the weights according to the degree of coincidence.

  According to an eighth aspect of the present invention, there is provided an imaging device, a drive unit capable of changing a distance between the imaging device and a table, an imaging surface of the imaging device, and a measured surface of a sample placed on the table. A scale for measuring a distance; a storage unit for storing an image photographed by the image sensor and an actual distance value measured by the scale; an imaging surface of the image sensor and the sample from the image and the actual distance value; A length-measuring method in a length-measuring device comprising a calculation unit for estimating a distance from a surface to be measured, wherein at least a part of the surface to be measured of the sample obtained by photographing the surface to be measured of the sample with the imaging device And two or more pieces of reference data including actual distance values measured on the scale when an image focused on at least a part of the surface to be measured of the sample is captured. Stored in the department Storing one or more target data including at least an image obtained by photographing the surface to be measured of the sample by the imaging device in the storage unit; an image of the reference data stored in the storage unit; A step of defining an arbitrary image processing range for the image of the target data; and for each image of the reference data and each image of the target data, each of the reference data A mean square value of a difference in gradation value in an overlapping region between the image and each image of the target data is calculated, and the minimum mean square value is obtained from the translation and rotation movements, and each image of the reference data is obtained. And a degree of coincidence between each image of the target data and an estimated distance value from the imaging element to the surface to be measured of the sample when each image of the target data is captured Measuring method characterized by comprising the steps of: estimating on the basis of the matching degree and the measured distance value of the corresponding reference data and the image of the reference data.

  According to the present embodiment, it is not necessary to focus on the entire area of the target sample, and it is not necessary to adjust the drive unit for focusing. Measurement is completed in a short time since the image is acquired by changing the distance to the sample by the drive unit, and the estimated distance of the sample is obtained with high accuracy based on the initial image and distance measurement value of the sample. be able to.

  Further, by performing a load test or the like and obtaining a measured value of the distance of the sample after the change with time, the position change of the uneven surface due to the change with time, that is, the change in the shape of the sample can be detected with high accuracy.

(A), (b) is a figure which shows the structure of the length measuring apparatus which concerns on Embodiment 1 of this invention. (A)-(c) is a figure which shows the image of the sample surface. It is a figure which shows the calculation process of the estimated distance in Embodiment 1 of this invention. It is a figure which shows the process process in the image process part of this embodiment. It is a figure which shows the calculation process of the deformation amount of the sample shape which concerns on Embodiment 2 of this invention. It is a figure which shows the structure of the length measuring apparatus which concerns on Embodiment 3 of this invention. It is a figure which shows the structure of the conventional length measuring apparatus. (A), (b) is a figure which shows the contrast of the image with respect to the movement amount of a Z direction.

  Hereinafter, embodiments of the present invention will be described in detail. ...

[Embodiment 1]
In FIG. 1, the structure of the length measuring apparatus which concerns on Embodiment 1 of this invention is shown. In FIG. 1, an apparatus including a light source 101, an image sensor 102, an optical element 103, a drive unit 104, and a scale 105 is shown. An optical element 103 such as a lens is appropriately disposed on the optical path between the light source 101 and the image sensor 102 and the sample 110. The light source 101, the image sensor 102, and the optical element 103 are connected to the table 106 via the drive unit 104.

  A sample 110 to be measured is mounted on a table 106, the surface thereof is brightened by light from the light source 101, and an image in which reflected light from the sample 110 is imaged is recorded by the image sensor 102. The light source 101 and the image sensor 102 can be driven in the z direction perpendicular to the table surface by a driving unit 104 fixed to the table 106, and the moved position is recorded by the scale 105. The position of the scale 105 indicates the distance between the imaging surface of the image sensor 102 and the surface of the table 106. If the thickness of the sample 110 on the table 106 does not change, the distance of the image surface of the image sensor 102 to the sample surface is determined. It shows relatively.

  The image obtained by the image sensor 102 and the position information measured by the scale 105 are not shown, but are recorded in the storage unit as electronic data and processed by the computer for image processing / estimation / comparison.

  The surface of the sample 110 is an uneven surface having a plurality of unevenness of about several μm to several tens of μm. A predetermined area on the uneven surface is recorded as an image, but since it is an uneven surface, the entire area in the region is not focused. The distance is adjusted so that at least a part of the region is in focus, and the distance at that time is set to z = z1.

  For example, as shown in FIG. 1A, when z = z1 is set so as to focus on the convex portion of the surface unevenness, an image as shown in FIG. 2A is obtained. FIG. 2A shows, for example, a digitized electronic file. A square of about 0.1 μm corresponds to one pixel, and 480 pixels × 600 pixels corresponds to a region of 48 μm × 60 μm. Corresponding to the pixel position information of the pixel, for example, it has a grayscale 8-bit gradation value. In FIG. 2A, it can be seen that the region A is not focused and the image is unclear, but the region B is focused and the image is clear.

  Next, the distance is adjusted so as to focus on other partial areas of the same image, and the distance at that time is set to z = z2. For example, as shown in FIG. 1B, when z = z2 is set so as to focus on the concave portion of the surface irregularities, an image as shown in FIG. 2B is obtained. Unlike FIG. 2A, in FIG. 2B, a clear image is obtained by focusing in the area A.

  At this time, the value of z = z1 and the image at that time, and similarly the value of z = z2 and the image at that time are recorded as a reference data set.

  Thereafter, the image is acquired by changing the distance so that at least a part of the same region is in focus, and this is used as a target data set. Images may be acquired continuously after acquiring the reference data set at z = z1 and z = z2, or once the sample is removed from the table 106 and another measurement or environmental load test is performed, the table is acquired again. It may be mounted on top to acquire images. Further, an image may be acquired by a measuring device that includes a light source, an imaging device, and a drive unit, and does not include the scale 105. Since at least a part is in focus within the same region, the distance z = Z′3 at that time is z1> Z′3> z2.

At this time, an image shown in FIG. 2C is obtained. Thereafter, the value of Z′3 is estimated from the degree of coincidence of the images. (Hereafter, “'” will be added to the estimated value.)
Here, a method for calculating the degree of coincidence will be described. First, the difference between the gradation values of pixels corresponding to pixel positions between two images is squared, and the result is summed over all the pixels, and then the mean square value obtained by dividing by the number of pixels is obtained. . Then, one image is translated or rotated with respect to the other image, the minimum value of the mean square value is obtained, and the obtained minimum mean square value is set as the degree of coincidence.

  The degree of coincidence between the image of the target data set and one of the images of the reference data set is calculated, and the smaller the mean square value, the more the images match. For example, the degree of coincidence C12 between the image of the target data set in FIG. 2C and the image of FIG. 2A as one of the reference data sets is 59. On the other hand, the degree of coincidence C22 between the image FIG. 2C and the image FIG. From this, it can be estimated that Z′3 is closer to z1 than z2 and Z′3≈z1. Further, for example, the distance may be weighted by the degree of coincidence, and Z′3 = (C1 × z2 + C2 × z1) / (C1 + C2) or Z′3 = (C12 × z2 + C22 × z1) / (C12 + C22) may be estimated. .

  These calculations need not be performed simultaneously with the measurement if the image is stored as an electronic file, and can be performed separately. In the above description, there are two reference data sets and one target data set. However, there are N reference data sets (N ≧ 2) and M target data sets (M ≧ 1). Also good.

  As described above, it is possible to estimate the distance at which the image is captured from the image. Due to the uneven surface, the conventional technique cannot focus on the entire region, and the measured distance varies between z2 and z1 each time measurement is performed. In addition, it may be focused locally with a laser or the like and scanned in the horizontal direction of the concave and convex surface, but because the drive unit is adjusted in order to focus at each point, the accuracy drops when applied to the entire concave and convex surface, There was a problem of taking time. In addition, when the measurement takes time, there is a problem that the measurement apparatus and the sample drift or are thermally deformed during the measurement, and the accuracy is further deteriorated.

  However, according to the present embodiment, it is not necessary to focus on the entire region, and it is sufficient to focus on a part of the region, and it is not necessary to adjust the drive unit for focusing. Since only the image is acquired by changing the distance regardless of the presence or absence of in-focus, it is possible to obtain an effect that an accurate value of the distance of the sample can be estimated by a short-time measurement.

  Since measurement can be performed in a short time, the scale is not deformed due to drift or thermal deformation of the measuring device, and accuracy is not deteriorated. In addition, the relationship between the distance between the image sensor and the stage in the reference data set can be replaced with the relative positional relationship of the height inherent to the sample, and further, the sample can be correlated with the image to make the sample independent of the measuring device. It can be a unique data set. Therefore, the target data set can be acquired even by another measuring apparatus that does not include a scale and includes a light source, an imaging device, and a drive unit, and thus the distance can be estimated.

  In FIG. 3, the calculation process of the estimated distance in Embodiment 1 of this invention is shown. This process mainly includes a process performed by the optical system of the light source 101, the image sensor 102, the optical element 103, the driving unit 104, and the scale 105, and a process performed mainly by a computer including a processing unit and a storage unit.

  First, the sample 110 is set on the stage 106 (S301), the drive unit 104 is adjusted to detect a partial range of the uneven surface that is focused, and the position of the scale 105 at that time is read. For the position, the measured distance zn from the imaging surface of the image sensor 102 to the uneven surface on which the sample 110 is focused and the corresponding image in are output as reference data to the storage unit (S302). S303 is repeated until the reference data reaches a desired number N (S303). When the lower limit and the upper limit distance of the adjustment range of the drive unit 104 are respectively za and zb, the reference data set has a distance za ≦ z1 <z2 <. . . <Zn <. . . N measured distances satisfying <zN ≦ zb, and images i1, i2,. . . in,. . . iN.

  When the reference data has been acquired to the desired number, a load test is performed on the sample 110 as necessary, and then the sample 110 is re-installed on the table 106 (S304). The drive unit 104 is adjusted to change the distance so that at least a part is in focus within the same region, and the captured image Im is output to the storage unit as target data (S305). S306 is repeated until the target data reaches the desired number M (S306).

  When the target data has been acquired to the desired number, the image processing unit calculates the degree of coincidence Cnm with all the images in (n = 1 to N) of the reference data set for the image Im (see FIG. S307). Among the N matching degrees C1m to CNm related to the image Im, the minimum matching degree Cnm_min is obtained, and the estimation unit estimates the estimated distance Z′m (S308). The estimated distance Z′m may be estimated by setting Z′m≈zn_min or weighting the distance zn with the matching degree Cnm as in the case of two.

  FIG. 4 shows processing steps in the image processing unit of this embodiment. The image processing unit includes an input unit, a conversion unit, an extraction unit, a matching degree calculation unit, a determination unit, and an output unit. The input unit designates an area for image processing in the image (S401). The conversion unit translates and rotates the designated area of the image (S402). The extraction unit superimposes two images of reference data and target data, and extracts an overlapping area (S403). The coincidence calculation unit calculates the coincidence from the gradation difference between the two images (S404). For example, the square of the gradation difference at the corresponding pixel position of the two images is obtained and averaged within the area designated in S401. The determination unit determines whether the execution has been completed within the image processing range in which the degree of coincidence is defined (S405). If completed, the degree of coincidence and the amount of translation and rotation when the degree of coincidence is minimized at the output unit are output to the estimation unit (S 406).

[Embodiment 2]
The second embodiment is different from the first embodiment in that an actually measured distance is acquired in addition to an image as a target data set. Even in the case of similar images, the thickness and shape of the sample may change due to changes over time, thermal expansion / contraction, swelling, etc., and the measured distance may change.

  For example, as in FIG. 2, it is assumed that FIG. 2 (a) is obtained when reference data is set and z = z1, and FIG. 2 (b) is obtained when z = z2, and measurement of target data is performed after the sample is subjected to an environmental load test. And the same image as FIG. 2A is obtained with z = Z3 as the target data set. In this case, since the images in FIG. 2A are the same, the surface is the same, but the measured distance z is different. Therefore, it is considered that Z3-z1 is the amount by which the position of the uneven surface is displaced by the environmental load test. Thus, when measuring the target data, the amount of deformation of the shape such as the thickness of the sample 110 can be detected by actually measuring the distance.

  In general, since the same image is rarely obtained as shown in FIG. 2A, the distance Z′3 at which the image of the target data set is obtained is estimated from the degree of coincidence as in the first embodiment. For example, Z′3 = (C1 × z2 + C2 × z1) / (C1 + C2). From this, the amount of change can be obtained by obtaining Z3-Z'3. The reference data set may include N reference data (N ≧ 2).

  The reference data set has the same configuration as that of the first embodiment, but the target data set includes M (M ≧ 1) images I1, I2,. . . , Im,. . . In addition to IM, the corresponding measured distances Z1, Z2, Z3,. . . , Zm,. . . Includes ZM. The estimated distance Z′m is estimated as in the first embodiment, and the amount of change of the sample 110 is estimated from the difference between the estimated distance Z′m and the measured distance Zm.

  The amount of change of the sample 110 is, for example, the reference data sets z1, z2,. . . , Zn,. . . , ZN can be estimated with an accuracy of at least Δz. Moreover, since the estimated value can be calculated by the number of target data (M), the amount of change is increased by increasing the number of target data, increasing the estimated value, and averaging the amount of change calculated based on each estimated value. The estimation accuracy of can be improved.

  Also, z1, z2,. . . zn,. . . zN may have different distance intervals. In this case, for any m, the minimum matching degree Cnm_min and the corresponding distance zn_min are obtained to obtain the estimated value Z′m, and the measured values (Z1, Z2,..., Zm,. , ZM) correspond to the estimated values (Z′1, Z′2,..., Z′m,..., Z′M) from the image, m) The constant c is estimated by regression analysis so that + c. This constant c is the amount of deformation of the sample shape. The estimation accuracy can be improved by performing regression analysis using M pieces of data.

  FIG. 5 shows a calculation process of the deformation amount of the sample shape according to the second embodiment of the present invention. The difference from Embodiment 1 is that the sample 110 is re-installed on the table 106 after the load test (S504), and then in S505, the measured distance Zm is recorded in the storage unit in addition to the image Im as target data, and in S509. In the newly added comparison unit, the difference between the estimated distance Z′m and the actually measured distance Zm is calculated and the displacement of the position of the uneven surface is calculated. S501 to S504, S507, and S508 are the same as S301 to S305, S307, and S308 of the first embodiment.

  Thus, in this embodiment, since the distance is also measured as the target data set, it is possible to estimate the amount of change over time of the uneven surface of the sample 110 that occurs between the acquisition of the reference data set and the acquisition of the target data set. It becomes.

[Embodiment 3]
FIG. 6 shows the configuration of the length measuring apparatus according to the third embodiment of the present invention. In this embodiment, the zero reference of the measured distance or the estimated distance can be taken as the reference plane.

  In FIG. 6, a sample 110 on a table 106 is provided with a reference 112 having a flat surface in addition to a measurement object 111 having an uneven surface, and the scale 105 uses the flat surface of the reference 112 as a zero reference. It is assumed that the position of the flat surface of the reference 112 does not change even when the position of the concavo-convex surface of the measurement object 111 changes due to an environmental load test or the like. If there is a pattern on the flat surface of the reference 112, the position of the flat surface of the reference 112 is measured by adjusting the positions of the light source 101 and the image sensor 102 to a position that maximizes the contrast of the image on the flat surface. Can do.

  Even if there is no pattern on the flat surface of the reference 112, it is equipped with a laser beam and a photodetector and adjusted to a position that maximizes the amount of light that is reflected by the flat surface of the reference 112 and returned. By doing so, the position of the flat surface of the reference 112 can be measured.

  In the first and second embodiments, the zero reference is on the apparatus side, but in the present embodiment, it is on the sample side. Since the apparatus is composed of complicated and different materials and mechanisms such as the light source 101 and the image sensor 102 such as the drive unit 104, the apparatus is more susceptible to environmental and temporal changes than the sample 110 having the reference 112. By providing the sample 110 with the reference 112 having a smaller change with time than the apparatus, it is possible to perform measurement with higher accuracy.

DESCRIPTION OF SYMBOLS 100 Length measuring apparatus 101 Light source 102 Image pick-up element 103 Optical element 104 Drive part 105 Scale 106 units 110 Sample 111 Measuring object 112 Reference | standard 700 Length measuring apparatus 701 Light source 702 Sensor 703 Drive part 704 Scale 710 Sample

Claims (8)

An image sensor;
A drive unit capable of changing a distance between the imaging element and the table;
A scale for measuring the distance between the imaging surface of the imaging device and the surface to be measured of the sample placed on the table;
An image focused on at least a part of the surface to be measured of the sample, and at least a part of the surface to be measured of the sample, taken by the imaging device on the surface to be measured of the sample placed on the table One or more reference data including at least two pieces of reference data including actual distance values measured at the scale when a focused image is photographed, and an image obtained by photographing the measured surface of the sample by the image sensor. A storage unit for storing target data;
With respect to an arbitrary image processing range of each image of the reference data and each image of the target data stored in the storage unit, each image of the reference data and each image of the target data are translated and rotated. Calculating the mean square of the difference between the gradation values in the overlapping region, and the minimum mean square among the translational and rotational movements as the degree of coincidence between each image of the reference data and each image of the target data, Based on the degree of coincidence with each image of the reference data and the corresponding measured distance value of the reference data, the estimated distance value from the image sensor to the measured surface of the sample when each image of the target data is captured Estimating, calculating part,
A length measuring device comprising: The target data includes an actual distance value measured on the scale when each image of the target data is taken,
2. The length measurement according to claim 1, wherein the calculation unit calculates a displacement of a position of the surface to be measured of the sample as a difference between the estimated distance value and a measured distance value of the corresponding target data. apparatus.   The calculation unit calculates the displacement of the position of the surface to be measured of the sample, and the actual distance value of the target data is the sum of the estimated distance value of the target data and the displacement of the position of the surface to be measured of the sample. The length measuring device according to claim 2, wherein the length measurement device calculates by performing regression analysis assuming that   When the range of the measured distance value of the reference data is z1 to z2, each image of the target data is taken at a position where the distance measured by the scale is between z1 and z2. The length measuring device according to any one of claims 1 to 3. The sample includes a reference plane different from the surface to be measured;
The length measuring apparatus according to claim 1, wherein the scale calculates a distance between an imaging surface of the imaging element and a surface to be measured of the sample with the reference surface as a zero reference.   The said calculation part makes the said estimated distance value of each said object data the said measured distance value of the said reference data which has an image with the said smallest coincidence degree. Length measuring device.   6. The calculation unit according to claim 1, wherein the calculation unit calculates the estimated distance value of each of the target data by weighting and adding the actually measured distance value of each of the reference data according to the degree of coincidence. The length measuring device according to any one of the above. An imaging device; a drive unit capable of changing a distance between the imaging device and the table; a scale for measuring a distance between an imaging surface of the imaging device and a surface to be measured of a sample placed on the table; and the imaging A storage unit that stores an image photographed by the element and an actual distance value measured by the scale, and a distance between the imaging surface of the image sensor and the measured surface of the sample is estimated from the image and the actual distance value. A length measuring method in a length measuring device comprising a calculation unit,
An image focused on at least a part of the surface to be measured of the sample and an image focused on at least a part of the surface to be measured of the sample obtained by photographing the surface to be measured of the sample with the imaging device. Storing in the storage unit two or more reference data including actual distance values measured at the scale when the image was taken;
Storing at least one target data including at least an image obtained by photographing the surface of the sample by the image sensor in the storage unit;
Defining an arbitrary image processing range for the image of the reference data and the image of the target data stored in the storage unit;
For each image of the reference data and each image of the target data, a gradation value in an overlapping region between each image of the reference data and each image of the target data is translated and rotated in the image processing range. Calculating a mean square value of the differences, obtaining a minimum mean square value from the translational and rotational movements, and obtaining a degree of coincidence between each image of the reference data and each image of the target data;
Based on the degree of coincidence with each image of the reference data and the corresponding measured distance value of the reference data, the estimated distance value from the image sensor to the measured surface of the sample when each image of the target data is captured Estimating, and
A length measuring method characterized by comprising: