JP2013096859A - Height measuring apparatus and height measuring method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology with which the height of each position of a sample having a mirror surface property can be determined accurately.SOLUTION: An imaging section 2 captures a sample image that is an image of a sample S, for example, at a predetermined frame rate. A spectrum extraction section extracts spectrums from sample images captured sequentially by the imaging section 2 and determines deviation w(x) of the spectrums with respect to a reference spectrum indicating predetermined reference height and reference inclination. A height calculation section sequentially calculates a height d(x) of the sample S from the reference height by substituting the deviation w(x) of the spectrums extracted by the spectrum extraction section to w(x)=2L×(d/dx)×d(x)+2sinθ×d(x).

Description

  The present invention is a technique for measuring the shape distribution of a sample, and particularly relates to a description for measuring unevenness of a semiconductor substrate, unevenness due to distortion of a rolled material, and the like.

  In recent years, a technology that irradiates the top surface of a mounted component with a light beam, projects specularly reflected light from the mounted component onto a screen, observes spot light, and measures the height of the mounted component based on changes in the position of the spot light Is disclosed (for example, Patent Document 1). And in patent document 1, when the measured height does not correspond with a reference | standard height, the technique which shifts a measurement position using the nominal value of the height of a mounted component is disclosed.

  However, in the method of Patent Document 1, it is necessary to previously acquire the nominal value of the height of the mounted component, and there is a problem that the height cannot be measured for a component whose nominal value is unknown.

In Patent Document 2, a virtual image of a rod-shaped light source reflected on a plate material is divided into blocks defined by X and Y coordinates, binarization processing is performed, and the form of bright lines is determined based on data of each divided area of bright lines. adaptively set the function expression ^{(Y = a0 + a1X + a2X} 2 + a3X 3 + a4X 4 + a5X 5 + a6X 6) to. A technique for improving the detection accuracy of the bright line by obtaining the X-direction position of each divided region of the bright line using this functional formula and approximating the bright line with the functional formula is disclosed.

  However, Patent Document 2 has a problem that when the sample has irregularities, the virtual image of the rod-shaped light source is bent and observed, and the irregularities of the sample cannot be detected with high accuracy.

  Patent Document 3 discloses the following technique. First, the specular reflection light of the aluminum rolled plate being conveyed is continuously imaged by a camera, and frame image data including a high-intensity portion showing the unevenness of the aluminum rolled plate is sequentially acquired. In this frame image data, the unevenness of the rolled plate is indicated by the Y coordinate, and the width direction of the rolled plate is indicated by the X coordinate. Then, for each X coordinate, the difference in the Y coordinate of the high luminance portion in a plurality of continuous frame image data is obtained, the maximum difference in the Y coordinate of the high luminance portion is specified, and the specified maximum difference is greater than a predetermined value In this case, a technique for determining that there is unevenness on one line of the corresponding X coordinate is disclosed.

JP 2010-140932 A JP-A-6-94439 JP 2010-139455 A

  However, in the method of Patent Document 3, the X coordinate whose height change is a certain level or more is simply specified, and it is determined whether or not there is unevenness on the X coordinate. It does not determine the height of each position.

  The objective of this technique is to provide the technique which can obtain | require accurately the height of each position of the sample which has specularity.

  (1) A height measuring apparatus according to the present invention receives a specularly reflected light from a sample irradiated by the light source, a light source that irradiates the sample with a diffused light, a conveyance unit that conveys the sample having specularity. An imaging unit that sequentially captures a sample image including a bright line that indicates the specular reflection light, and a reference bright line that extracts a bright line from the sample image sequentially captured by the imaging unit and indicates a predetermined reference height and reference inclination A bright line extraction unit for obtaining a shift of each bright line with respect to the height, a height shift component indicating a shift of the bright line from the reference bright line based on the height of the sample from the reference height, and the sample from the reference tilt Substituting the deviation of the bright line extracted by the bright line extraction unit into an equation representing the deviation of the bright line with respect to the reference bright line by a linear combination with an inclination shift component indicating the deviation of the bright line from the reference bright line based on the tilt. , And a height calculation unit for sequentially calculating the height of the sample.

  According to this structure, the sample image containing the bright line which shows regular reflection light is sequentially acquired by imaging continuously the sample currently conveyed. And a height deviation component indicating deviation of the bright line from the reference bright line based on the height of the sample from the reference height, and an inclination deviation component indicating deviation of the bright line from the reference bright line based on the inclination of the sample from the reference inclination. The height of the sample is obtained by using an equation representing the deviation of the bright line from the reference bright line by the linear combination of. Therefore, the height of the sample having specularity can be obtained with high accuracy.

  Further, since the sample being transported is continuously imaged, the height of each position in the sample transport direction can be determined sequentially, and the surface shape of the sample can be determined.

  (2) The height deviation component is represented by 2 sin θ · d (x), the inclination deviation component is represented by 2L · (d / dx) · d (x), and the equation is expressed as w = 2L · (D / dx) · d (x) +2 sin θ · d (x) is preferable.

  Where x represents the measurement position of the sample with respect to the transport direction, w (x) represents the deviation of the bright line from the reference bright line, L represents the optical path length of the specular reflected light, and d (x) represents the above The height of the sample from the reference height is indicated, (d / dx) · d (x) indicates the inclination of the sample from the reference inclination, and θ indicates the reflection angle of the specular reflection light.

  According to this configuration, the height d (x) at the position x can be obtained by substituting the deviation w (x) of the bright line with respect to the reference bright line and solving the equation for d (x).

  (3) It is preferable that the sample has streak-like unevenness, and the light source is a rod-shaped light source arranged so that a longitudinal direction thereof is perpendicular to a longitudinal direction of the unevenness of the sample.

  According to this configuration, since the longitudinal direction of the rod-shaped light source is arranged so as to be orthogonal to the longitudinal direction of the unevenness of the sample, for example, a specularity is applied to a sample having streaky unevenness such as a solar cell wafer or an aluminum rolled material. Can be given.

  (4) The reference bright line is preferably the bright line of the sample at the initial measurement position.

  According to this configuration, since the bright line at the initial position is set as the reference bright line, the reference bright line can be obtained without using the reference sample different from the sample and measuring the bright line at the initial position.

  (5) A laser light source for irradiating the sample with a light cutting line, and a light cutting line irradiated by the laser light source is imaged by the imaging unit, and the width of the sample is obtained based on the obtained light cutting line image A reference height calculation unit that calculates the height of each initial measurement position in the direction, and the height calculation unit uses the height of each initial measurement position calculated by the reference height calculation unit. It is preferable to calculate the measurement position of the height of each position of the sample.

  According to this configuration, an initial value required when solving the equation shown in (1) can be obtained, and the height of the sample can be obtained with high accuracy. In addition, the height of each initial measurement position in the width direction of the sample can be obtained, and thereafter, the height of each position in the transport direction with respect to each position in the width direction can be obtained using this initial height. The height of the sample can be calculated in consideration of the relative height difference in the width direction of the sample.

  According to the present invention, sample images including bright lines indicating specularly reflected light are sequentially acquired by continuously imaging the sample being conveyed. And a height deviation component indicating deviation of the bright line from the reference bright line based on the height of the sample from the reference height, and an inclination deviation component indicating deviation of the bright line from the reference bright line based on the inclination of the sample from the reference inclination. The height of the sample is obtained by using an equation representing the deviation of the bright line from the reference bright line by the linear combination of. Therefore, the height of the sample having specularity can be obtained with high accuracy.

  Further, since the sample being transported is continuously imaged, the height of each position in the sample transport direction can be obtained sequentially, and data indicating the surface shape of the sample can be obtained.

It is a whole lineblock diagram of the height measuring device by an embodiment of the invention. It is the figure which showed an example of a sample image, (A) has shown the sample image when a sample is flat, (B) has shown the sample image when a sample has an unevenness | corrugation. It is a figure for demonstrating the principle of this invention, and is the figure which looked at the sample from the j-axis direction in FIG. It is a block diagram of the height measuring apparatus shown in FIG. It is a whole block diagram of the height measurement apparatus in the case of calculating initial height using a light cutting method. It is the figure which showed the positional relationship of the laser light source 5 and the sample S in the x-axis direction view. It is a flowchart which shows operation | movement of the height measuring apparatus by embodiment of this invention.

  FIG. 1 is an overall configuration diagram of a height measuring apparatus according to an embodiment of the present invention. As shown in FIG. 1, the height measuring device includes a light source 1, an imaging unit 2, and a transport unit 3. The sample S is made of a plate material having specularity with streaky irregularities.

  Further, as the sample S, a flat plate-like object having periodic streak irregularities can be adopted. Specifically, as the sample S, a solar cell wafer obtained by cutting a cylindrical ingot in a direction orthogonal to the longitudinal direction, or a rolled material obtained by rolling aluminum can be employed. When the rolled material is rolled, a streak-shaped rolling bar is formed in a certain direction. Further, when a solar cell wafer is cut into an ingot, streaky irregularities called saw marks are formed in a certain direction. In the present embodiment, the sample S having such streak-like irregularities is the measurement target. The sample S is arranged so that the longitudinal direction of the streaky irregularities is substantially parallel to the transport direction A.

  In the following description, the y-axis is set in the height direction orthogonal to the main surface of the sample S, the x-axis is set in the direction parallel to the transport direction A, and the sample S orthogonal to the x direction and the y direction is set. Set the j-axis in the width direction.

  The light source 1 is composed of a rod-shaped light source arranged so that the longitudinal direction is perpendicular to the transport direction A of the sample S when the sample S is viewed from directly above, that is, the j-axis direction. On the other hand, diffuse light is irradiated. Specifically, a fluorescent lamp is adopted as the light source 1. Here, as the length of the light source 1 in the longitudinal direction, for example, it is preferable to have a length capable of irradiating diffused light to the entire region in the width direction of the sample S. It is preferable to have a length longer than the width direction.

  The imaging unit 2 includes a moving image camera that captures a sample image that is an image of the sample S at a predetermined frame rate, for example. Here, the imaging unit 2 receives the specular reflection light L2 of the light L1 emitted from the light source 1. The imaging unit 2 includes a lens and a light receiving element (not shown), and the light receiving element receives the specular reflection light L2 via the lens.

  Here, as shown in FIG. 3, the light source 1 and the imaging unit 2 have both a distance x between the measurement position in the x-axis direction and the distance between the light sources 1 and a distance between the position x and the imaging unit 2. It arrange | positions so that it may become L. Further, the angle formed between the normal line H1 of the sample S and the imaging unit 2 at the position x and the angle formed between the normal line H1 and the light source 1 are both θ and equal.

  Therefore, the imaging unit 2 can obtain a sample image in which the specular reflection light L2 at the position x of the light L1 is represented as a bright line, that is, a sample image in which the mirror image 1 ′ of the light source 1 with respect to the sample S is represented as a bright line. .

  In the present embodiment, each measurement position is defined by a plurality of coordinates arranged at regular intervals on the j-axis and a plurality of coordinates arranged at regular intervals on the x-axis. That is, the measurement positions are distributed in a matrix on the jx plane. Therefore, the height of the entire main surface of the sample S can be obtained.

  Returning to FIG. 1, the transport unit 3 includes a plurality of transport rollers 31. In FIG. 1, three transport rollers 31 are shown, but this is an example, and may be 4, 5, 6 or more, in order to transport the sample S along the transport direction A without rattling. Further, it is preferable that the interval between the transport rollers 31 is shorter than the length of the sample S in the x direction.

  The sample S is placed on the upper side of the transport roller 31, and the transport roller 31 is rotationally driven by a motor (not shown) to transport the sample S. Here, the transport unit 3 transports the sample S at a constant speed in the transport direction A. The sample S is imaged at a constant frame rate by the imaging unit 2 while being conveyed at a constant speed in the conveyance direction A. Therefore, sample images are sequentially acquired at a constant cycle.

  2A and 2B are diagrams showing an example of a sample image. FIG. 2A shows a sample image G1 when the sample S is flat, and FIG. 2B shows a sample image G1 when the sample S has irregularities. Yes.

  2A and 2B, the i-axis indicates the vertical coordinate of the sample image G1, and the j-axis indicates the horizontal coordinate of the sample image G1. The j axis corresponds to the j axis shown in FIG. 1 and indicates the width direction of the sample S. In FIG. 2A, the shape of the sample S in the j-axis direction is parallel to the ground surface and is flat. Therefore, the bright line KL appears in parallel with the j-axis in the sample image G1 in FIG.

  On the other hand, in the sample image G1 in FIG. 2B, the bright line KL protrudes near the center of the j-axis. Therefore, in the sample image G1 in FIG. 2B, the i-axis value gradually increases toward the center of the j-axis, and has a convex shape.

  FIG. 4 is a block diagram of the height measuring apparatus shown in FIG. As shown in FIG. 4, the height measuring device includes a light source 1, an imaging unit 2, a transport unit 3, a control unit 4, and a laser light source 5. The control unit 4 includes, for example, a microcontroller, and includes a conveyance control unit 41, a light source control unit 42, a bright line extraction unit 43, a height calculation unit 44, and a reference height calculation unit 45.

  The conveyance control unit 41 controls the conveyance unit 3. For example, the conveyance control unit 41 outputs a control signal to a motor that drives the conveyance roller 31, and drives the motor to drive the conveyance roller 31. Here, the conveyance control unit 41 starts driving of the conveyance unit 3 by pressing a measurement start button by an operator, for example.

  The light source control unit 42 supplies a drive signal to the light source 1 to turn on the light source 1. For example, when the measurement start button is pressed, the light source 1 is turned on.

  The bright line extraction unit 43 extracts the bright line KL from the sample image G1 sequentially picked up by the image pickup unit 2, and obtains the deviation of each bright line KL from the reference bright line KL_0 indicating a predetermined reference height and reference inclination. Here, for example, as shown in FIG. 2A, the bright line extraction unit 43 sets a target line CL (j) parallel to the i-axis for the sample image G1, and on the target line CL (j). The pixel gmax (j) having the highest luminance is searched. When the search for the target line CL (j) is completed, j is incremented by 1, and the pixel gmax (j) having the highest luminance in the next target line CL (j) is searched. Then, this search is repeatedly executed for each target line CL (i) from j = 0 to j = J while increasing j by 1. Then, a pixel group composed of gmax (0) to gmax (J) is extracted as the bright line KL. That is, the bright line extraction unit 43 extracts one bright line KL from one sample image G1. J indicates the maximum value of the measurement position of the sample S on the j-axis.

  In this case, the pixel gmax (j) having the highest luminance may be searched for in subpixel units. Specifically, the luminance curve of the target line CL (j) is obtained by interpolating the luminance of each pixel arranged in the target line CL (j) using an interpolation process such as spline interpolation. For example, the i coordinate of the center of gravity may be obtained up to the first decimal place and searched as the pixel gmax (j).

  Then, the bright line extraction unit 43 obtains the deviation of the bright line KL from the reference bright line KL_0. As shown in FIG. 3, the sample S before the upper sample S is shifted is shown, and the sample S after the lower sample S is shifted downward by d (x) along the y axis is shown. A bright line KL that appears in the sample image when the sample S before being irradiated with the light L1 is defined as a reference bright line KL_0. In the present embodiment, the bright line KL at the initial measurement position where x = 0 is adopted as the reference bright line KL_0. The reference height indicates the height of the sample S with x = 0, and the reference inclination indicates the inclination of the sample S with x = 0.

  Then, the bright line extraction unit 43 sets i_KL_0 (j) as the i-axis value at j (= 0 to J) with the reference bright line KL_0, and i_KL (j) as the i-axis value at j with the bright line KL. i_KL (j) -i_KL_0 (j) is obtained as a shift w (x) of the bright line KL with respect to the reference bright line KL_0. In other words, the bright line extraction unit 43 obtains the deviation of the bright line KL from the reference bright line KL_0 by taking the difference of the i-axis values with the same j in the reference bright line KL_0 and the bright line KL. Thereby, the height calculation unit 44 obtains the height d (x) at each position in the width direction defined by j.

  The height calculation unit 44 includes a height deviation component indicating a deviation of the bright line KL from the reference bright line KL_0 based on the height of the sample S from the reference height, and a reference bright line KL_0 based on the inclination of the sample S from the reference inclination. By substituting the deviation of the bright line KL extracted by the bright line extraction unit 43 into an equation representing the deviation of the bright line KL with respect to the reference bright line KL_0 by linear combination with the inclination deviation component indicating the deviation of the bright line KL of the sample S, The length d (x) is calculated sequentially.

  Specifically, the height deviation component is represented by 2 sin θ · d (x), the inclination deviation component is represented by 2L · φ (x), and the equation is represented by Expression (1).

w (x) = 2L · φ (x) +2 sin θ · d (x) (1)
Where w (x) indicates the deviation of the bright line KL from the reference bright line KL_0, L indicates the optical path length of the specular reflected light L2, d (x) indicates the height of the sample S from the reference height, and φ (X) indicates the inclination of the sample S from the reference inclination, and θ indicates the reflection angle of the specular reflection light L2.

  FIG. 3 is a view for explaining the principle of the present invention, and is a view of the sample S viewed from the j-axis direction in FIG. When the sample S specularly reflects the light L1, in the imaging unit 2, the mirror image 1 'of the light source 1 appears as a bright line KL in the sample image G1.

  Hereinafter, assuming that the sample S is flat, the deviation w (x) of the bright line KL, the inclination φ (x) of the sample S from the reference inclination, and the height d (x) of the sample S from the reference height The relationship will be described.

  The mirror image 1 ′ is located at a position where the light source 1 is folded back symmetrically with respect to the sample S. When the sample S changes down by a distance d (x) along the y axis, the mirror image 1 'is observed as if it moved down by 2d (x) along the y axis. Therefore, the change of the mirror image 1 ′ in the direction orthogonal to the specular reflection light L <b> 2 can be expressed as 2 sin θ · d. Therefore, the deviation w (x) of the mirror image 1 ′ observed by the imaging unit 2 when the sample S changes along the y axis by the distance d (x), that is, the height deviation component is 2 sin θ · d. Become.

  On the other hand, if the sample S is tilted by φ around the position x, the mirror image 1 ′ is observed as if it is rotated approximately 2L · φ in the imaging unit 2 when φ is small. Therefore, the deviation of the mirror image 1 ′ in the direction orthogonal to the specular reflection light L <b> 2 can be expressed as 2L · φ. Therefore, the tilt deviation component is represented by 2L · φ.

  As described above, the shift w (x) of the bright line KL at the position x is expressed by a linear combination of the height shift component and the tilt shift component. Therefore, if the initial value that is the value of the height d (0) at x = 0 is known, the height d (x) of the sample S relative to the reference height from the deviation w (x) of each bright line KL with respect to the reference bright line KL_0. ) Can be calculated sequentially.

  Here, since the sample S moves in the x direction, the value of the deviation w (x) of the bright line KL changes due to the influence of the unevenness. Therefore, if the deviation w (x) of the bright line KL is known, the height d (x) of the sample S from the reference height can be obtained.

  On the other hand, the inclination φ (x) is also a function of x, but since this value is a change in the inclination of the sample S, it coincides with the differential value of d (x) in the x direction. That is, the formula (1) is expressed as the following formula (2).

d (x) = 2L · (d / dx) · d (x) +2 sin θ · d (x) (2)
Equation (2) is a differential equation, and is expressed as equation (3) when d is the integral constant and C is solved for d (x).

d (x) = exp (-sinθ ・ x / L) ∫ (w (x) / 2L) ・ exp (sinθ ・ x / L) dx + C （３）
Note that it is sufficient to know the relative change of the sample S in the measurement, and the value of C is not necessarily measured.

  In the present embodiment, the height calculation unit 44 sets d (0) = 0, and thereafter calculates d (1), d (2), d (3),..., D (x) sequentially. To do. Therefore, at x = 0, the height d (0) of each measurement position on the j-axis is all regarded as 0, and only a relative height change in the x-axis direction is observed.

  The solar cell wafer and the rolled material have streaky irregularities and have a rough surface. When the member having such surface roughness is the sample S, the light L1 is generally not specularly reflected. On the other hand, even if the sample S has a surface roughness, the light L1 can be specularly reflected as shown in FIG. 3 if the longitudinal direction of the unevenness is arranged so as to be orthogonal to the x-axis direction. Thereby, the sample S can be given a mirror surface property.

  As shown in FIG. 2A, when the sample S can be regarded as a plane, the mirror image 1 'observed by the imaging unit 2 is parallel to the j-axis. On the other hand, when the sample S has irregularities, the mirror image 1 ′ observed by the imaging unit 2 changes in the i-axis direction according to the irregularities. Therefore, if the deviation w (x) of the bright line KL shown in FIG. 2B with respect to the bright line KL shown in FIG. 2A is known, the height d (x) from the reference height can be known.

  And by calculating | requiring such height d (x) for every value on a j-axis, the height d (x) in each position of the whole area of the width direction of the sample S can be calculated | required.

  Equation (3) can be expressed by the product-sum calculation shown in Equation (4).

d (n, j) = (1 / 2sinθ) ・ w (n, j)-(L / sinθ) ・ (d / dx) d (n, j) (4)
Here, assuming that the measurement interval in the x-axis direction is Δx, the deviation of the mirror image 1 ′ in the nth frame is w (n, j), and the height is d (n, j), Equation (4) becomes Equation (5). ).

d (n, j) = (1 / 2sinθ) · w (n, j)-(L / sinθ) · ((d (n-1, j) -d (n-2, j)) / Δx) ( 5)
It should be noted that d (n, j) shown in equations (4) and (5) corresponds to d (x) in equation (2).

  When calculating d (n, j) sequentially using Equation (5), it is necessary to determine d (0, j) and d (1, j) in advance. In this embodiment, d (1, j) = d (0, j) = 0 is set, and d (2, j) = (1/2 sin θ) · w (2, j) is obtained. Thereafter, d (n, j) is obtained sequentially.

  In order to increase the accuracy, d (1, j) and d (0, j) are obtained using a light cutting method, and the obtained d (1, j) and d (0, j) are used. D (2, j) may be obtained.

  This is realized by the reference height calculation unit 45 shown in FIG. That is, the reference height calculation unit 45 causes the imaging unit 2 to capture the light cutting line irradiated by the laser light source 5 onto the sample S, and acquires a light cutting line image at n = 0, 1. Then, the reference height calculation unit 45 extracts the light cutting line included in the light cutting line image at each of n = 0 and 1. Then, the initial height of the sample S is calculated from d (0, j) and d (1, j) from the coordinates of the extracted light cutting line at n = 0,1.

  FIG. 5 is an overall configuration diagram of a height measuring apparatus in the case where the initial height is calculated using the light cutting method. In the height measuring apparatus shown in FIG. 5, a laser light source 5 is arranged with respect to the height measuring apparatus shown in FIG. The laser light source 5 includes, for example, a semiconductor laser and an optical system that expands the light output from the semiconductor laser in a fan shape, and irradiates the sample S with the optical cutting line JL.

  Here, the laser light source 5 irradiates the sample S with the light cutting line JL from above. Specifically, the laser light source 5 is arranged so that the elevation angle with respect to the sample S is 90 degrees.

  The imaging unit 2 images the sample S irradiated with the light cutting line JL, and obtains a light cutting line image. The reference height calculation unit 45 extracts the light section line JL included in the light section line image by using the same technique as that used to extract the bright line KL.

  Specifically, the reference height calculation unit 45 uses the i-coordinate value of the light cutting line JL at each position on the j-axis, the elevation angle of the laser light source 5, and the elevation angle of the imaging unit 2 to determine the j-axis. Find the height y for each position above.

  FIG. 6 is a diagram showing a positional relationship between the laser light source 5 and the sample S when viewed in the x-axis direction. As shown in FIG. 6, the laser light source 5 is provided on the normal line H <b> 1 passing through the center of the sample S in the width direction, and is disposed immediately above the sample S.

  Here, if the elevation angle of the imaging unit 2 with respect to the normal line H1 is α, and the elevation angle of the laser light source 5 with respect to the normal line H1 is β (not shown), the height for each position on the j-axis is expressed by Equation (6). Calculated.

y = k * i_JL (j) * cos [beta] / sin ([alpha] + [beta]) (6)
However, k represents the pixel resolution of the imaging unit 2, i_JL (j) represents the value of the i-axis at each position on the j-axis of the light section line JL, β represents the elevation angle of the laser light source 5, and α represents imaging. The elevation angle of part 2 is shown. Further, the pixel resolution k is known. 5 and 6, since the elevation angle β of the laser light source 5 is β = 0, the height y is obtained by y = k × g_JL (j) / sin (α).

  As shown in FIG. 6, the laser light source 5 and the imaging unit 2 do not satisfy the specular reflection condition. Therefore, the imaging unit 2 does not receive the specular reflection light of the laser light source 5, and it is difficult to observe the light section line JL. However, if the optical cutting line JL is imaged over time, an optical cutting line image in which the optical cutting line JL appears clearly can be obtained.

  Specifically, with the sample S fixed, the sample S is irradiated with the light cutting line JL for a certain period of time, and a light cutting line image at n = 0 is acquired. Similarly, a light section line image at n = 1 may be acquired with the sample S fixed.

  As described above, by adopting the height using the optical cutting method as the height d (0, j) and d (1, j) of the initial position, the relative difference in the height in the j-axis direction is also obtained. Taking into account, the height of the sample S can be observed.

  FIG. 7 is a flowchart showing the operation of the height measuring apparatus according to the embodiment of the present invention. First, the light source control unit 42 turns on the light source 1 when receiving an instruction to start measurement from the operator (S1). Next, the conveyance control part 41 starts conveyance of the sample S (S2). Thereby, the sample S is conveyed while being irradiated by the light source 1.

  Next, the bright line extraction unit 43 causes the imaging unit 2 to image the sample S and acquires a sample image (S3). Thereby, the sample image G1 with the bright lines KL as shown in FIGS. 2A and 2B is acquired.

  Next, the bright line extraction unit 43 extracts the bright line KL from the sample image G1 (S4). Next, the bright line extraction unit 43 calculates a shift w (n, j) of the bright line KL of the current frame with respect to the reference bright line KL_0 (S5). Note that the bright line KL with n = 0 is the reference bright line KL_0.

  Next, the height calculation unit 44 calculates the height d (n−2, j) of the sample S two frames before, the height d (n−1, j) of the sample S one frame before, and the current frame. The deviation w (n, j) of the bright line KL is substituted into the equation (5), and the height d (n, j) of the sample S in the current frame is calculated (S6). Note that the height calculation unit 44 may calculate the height d (n, j) with the height d (0, j) = d (1, j) = 0, or may be calculated using a light cutting method. The height d (n, j) may be calculated using the heights d (0, j) and d (1, j).

  Next, when the sample S is being transported and the measurement has not been completed (NO in S7), the height calculation unit 44 increases the variable n for specifying the frame by 1, and returns the process to S3.

  On the other hand, when the conveyance of the sample S is completed and the measurement is completed (YES in S7), the height calculation unit 44 ends the process.

  As described above, the height measuring apparatus according to the present embodiment is adapted to the deviation of the bright line KL based on the inclination φ (x) of the sample S and the height d (x) of the sample S as shown in the equation (1). The height d (x) is calculated in consideration of the deviation w (x) of the bright line KL based thereon. Therefore, the height of each position of the sample S can be obtained with high accuracy.

<Modification 1>
It is assumed that the unevenness of the sample S to be measured by the height measuring device according to the present embodiment changes with a limited spatial frequency in the longitudinal direction. Therefore, a filtering process is performed on the data group indicating the height d (n, j) of the sample S toward the x-axis to extract the assumed spatial frequency band height component d ′ (n, j). May be. Specifically, a one-row × predetermined one-dimensional filter for extracting data of an assumed spatial frequency band is converted into a data group composed of one column height d (n, j) along the x-axis. The height component d ′ (n, j) of the assumed spatial frequency band may be extracted by application. Thereby, the noise component contained in height d (d, n) can be removed.

1 'mirror image 1 light source 2 imaging unit 3 transport unit 4 control unit 5 laser light source 31 transport roller 41 transport control unit 42 light source control unit 43 bright line extraction unit 44 height calculation unit 45 reference height calculation unit A transport direction G1 sample image JL Light cutting line KL Bright line KL_0 Reference bright line L1 Light L2 Specular reflection light S Sample

Claims (6)

A transport unit for transporting a sample having specularity;
A light source for irradiating the sample with diffused light;
An imaging unit that receives specular reflection light from the sample irradiated by the light source and sequentially captures sample images including bright lines indicating the specular reflection light;
A bright line extraction unit that extracts a bright line from a sample image sequentially captured by the imaging unit and obtains a shift of each bright line with respect to a reference bright line indicating a predetermined reference height and reference inclination;
A height shift component indicating a shift of the bright line from the reference bright line based on the height of the sample from the reference height, and a shift of the bright line from the reference bright line based on the tilt of the sample from the reference tilt. The height of the sample is sequentially calculated by substituting the deviation of the bright line extracted by the bright line extraction unit into an equation representing the deviation of the bright line with respect to the reference bright line by a linear combination with an inclination deviation component indicating A height measurement apparatus comprising a height calculation unit. The height deviation component is represented by 2 sin θ · d (x),
The tilt deviation component is represented by 2L · (d / dx) · d (x),
The height measurement apparatus according to claim 1, wherein the equation is represented by w = 2L · (d / dx) · d (x) +2 sin θ · d (x).
Where x represents the measurement position of the sample with respect to the transport direction, w (x) represents the deviation of the bright line from the reference bright line, L represents the optical path length of the specular reflected light, and d (x) represents the above The height of the sample from the reference height is indicated, (d / dx) · d (x) indicates the inclination of the sample from the reference inclination, and θ indicates the reflection angle of the specular reflection light. The sample has streaky irregularities,
The height measuring apparatus according to claim 1, wherein the light source is a rod-shaped light source arranged so that a longitudinal direction thereof is perpendicular to a longitudinal direction of the unevenness of the sample.   The height measuring apparatus according to claim 1, wherein the reference bright line is a bright line of the sample at an initial measurement position. A laser light source for irradiating the sample with an optical cutting line;
A reference height calculation unit that images a light cutting line irradiated by the laser light source with the imaging unit and calculates the height of each initial measurement position in the width direction of the sample based on the obtained light cutting line image And further comprising
The height calculation unit calculates the measurement position of the height of each position of the sample using the initial height of each measurement position calculated by the reference height calculation unit. The height measuring device described in 1. A transport step for transporting a sample having specularity;
An irradiation step of irradiating the sample with diffused light;
An imaging step of receiving specular reflection light from the sample irradiated by the light source and sequentially capturing sample images including bright lines indicating the specular reflection light;
A bright line extraction step for extracting a bright line from the sample image sequentially captured by the imaging step and obtaining a shift of each bright line with respect to a reference bright line indicating a predetermined reference height and reference inclination;
A height shift component indicating a shift of the bright line from the reference bright line based on the height of the sample from the reference height, and a shift of the bright line from the reference bright line based on the tilt of the sample from the reference tilt. The height of the sample is sequentially calculated by substituting the deviation of the bright line extracted by the bright line extraction step into an equation representing the deviation of the bright line with respect to the reference bright line by a linear combination with an inclination deviation component indicating A height measurement method comprising a height calculation step.

Images