JP2012242138A - Shape measuring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To calculate at high speed not only shapes of front and rear surfaces of a solar battery wafer, but also a thickness of the solar battery wafer.SOLUTION: Light sources 121 and 131 irradiate front and rear surfaces of a solar battery wafer with a light cutting line CL. Cameras 122 and 132 perform continuous imaging of a light cutting line image of the front and rear surfaces of a measurement sample 500, every time the solar battery wafer is transported a predetermined distance. Measurement data calculating parts 123 and 133 calculate, as front surface measurement data and rear surface measurement data, gravity center coordinates where the light cutting line CL appears from an angle light cutting line image. A height data calculating part 143 calculates height data on the front and rear surfaces of the solar battery wafer from the front surface measurement data and the rear surface measurement data. A thickness data calculating part 146 obtains thickness data on the solar battery wafer from the height data of the front and rear surfaces of the solar battery wafer.

Description

  The present invention relates to a shape measuring apparatus that automatically measures a thickness, which is an appearance inspection item of a solar cell wafer, at high speed.

  Conventional visual inspection of solar cell wafers (abbreviated as “wafers”) has been carried out by humans, but there are various reasons such as the recent expansion of the solar cell market and reduction of wafer breakage caused by humans when inspecting solar cell wafers. Therefore, high-speed automatic inspection equipment is being studied. An apparatus has already been developed and put to practical use for measuring the unevenness of 1 to 3 lines on both sides of the wafer using a laser displacement meter and inspecting the thickness from the displacement difference. However, an apparatus for inspecting the entire thickness of the front and back surfaces of the wafer has not been developed.

  Patent Document 1 discloses an apparatus in which optical distance meters are arranged on both upper and lower sides of an object to be measured, and the thickness of the object to be measured is determined from the distance between the optical distance meters and the measured values of the two distance meters. Yes. Specifically, one optical distance meter is provided with two projectors and one imaging system, and irradiates two light beams so as to intersect two different light beams from different directions with respect to the surface to be measured. The spot interval ΔX and the center position Xc of the light spot are obtained. Then, using the fact that the position of the center position Xc changes when the surface to be measured is inclined by the angle θ, the angle θ is detected from the amount of change of the center position Xc. Then, the distance between the upper optical distance meter and the upper measured surface is obtained from the distance ΔX obtained by the upper optical distance meter, and the lower distance is obtained from the distance ΔX obtained by the lower optical distance meter. The distance between the optical distance meter and the lower measurement surface is obtained, and the measurement thickness t ′ of the measurement object is obtained. And it is disclosed that the measured thickness t ′ is corrected by t ′ × cos θ to determine the actual thickness t.

  Patent Document 2 discloses a technique for accurately obtaining the thickness of a measurement object in consideration of the tilt angle and the twist angle of the measurement object. Specifically, the heights of the upper and lower surfaces of the object to be measured are respectively measured with the upper contactor and the lower contactor. Then, a difference Δh between the height of the measurement object when measured at the position ha in the lower contact and the height when the measurement object is measured at the position hb (| ha−hb | = Δd) is obtained. Then, the inclination angle θ = arctan (Δh / Δd) of the object to be measured is obtained. And it is disclosed that the measured thickness t ′ obtained from the difference between the height of the upper contact and the height of the lower contact is corrected by the actual thickness t = t ′ × cos θ to obtain the actual thickness t. Yes.

JP 2006-189389 A JP 2006-30044 A

  However, the method of Patent Document 1 is limited to the case where the thickness of the object to be measured is constant, and cannot cope with the case where the thickness varies. Therefore, it is not possible to accurately calculate the thickness of the solar cell wafer having a different thickness depending on the position. Moreover, in patent document 1, since the cross-sectional shape of only one line is measured, the thickness of the whole surface of a solar cell wafer cannot be calculated, and distribution of the calculated thickness cannot be evaluated.

  In Patent Document 2, since the thickness is obtained by bringing the upper and lower contacts into contact with both surfaces of the object to be measured, the thickness of the entire surface of the solar cell wafer cannot be obtained at high speed.

  An object of the present invention is to provide a shape measuring device capable of calculating not only the shape of the front and back surfaces of a solar cell wafer but also the thickness of the solar cell wafer at high speed.

  (1) A shape measuring apparatus according to the present invention includes a transport unit that transports a solar cell wafer, and an optical cutting line in a direction that intersects the transport direction of the solar cell wafer from an oblique direction with respect to the front and back surfaces of the solar cell wafer. The measurement data acquisition unit that continuously images the front and back surfaces of the solar cell wafer and acquires the surface measurement data and the back surface measurement data of the solar cell wafer, and the surface measurement data acquired by the measurement data acquisition unit And a height data calculation unit for calculating height data at each position on the front surface of the solar cell wafer and height data at each position on the back surface based on the back surface measurement data, and calculated by the height data calculation unit. A thickness data calculation unit for calculating thickness data of each position of the solar cell wafer based on the height data of the front surface and the height data of the back surface of the solar cell wafer; That.

  According to this configuration, since the thickness of the solar cell wafer is measured using the height data of the front and back surfaces of the solar cell wafer, the thickness of the solar cell wafer as well as the shape of the front and back surfaces of the solar cell wafer is measured. Can also be obtained. Therefore, the quality of the solar cell wafer can be determined using the thickness in addition to the saw mark. Moreover, since the surface measurement data and the back surface measurement data are obtained by continuously imaging the front surface and the back surface of the solar cell wafer at the same time, the height data and the thickness data of the solar cell wafer can be obtained at high speed.

  (2) The height data calculation unit has the same thickness as the solar cell wafer, and the height data of the front and back surfaces acquired by the measurement data acquisition unit measuring a flat sample with flat front and back surfaces. Based on the height data of the front and back surfaces obtained by measuring a thickness sample having a known thickness, the standard surface height data and the standard back surface height data are calculated, and the standard surface height data and the standard back surface data are calculated. Based on the height data, a parameter calculation unit for calculating the front surface tilt correction parameter and the back surface tilt correction parameter, and the front surface height data and the back surface height data of the solar cell wafer calculated by the height data calculation unit. Is corrected using the front surface inclination correction parameter and the back surface inclination correction parameter, and the height data for calculating the corrected front surface height data and the corrected back surface height data is calculated. Anda data correction unit, the thickness data calculating unit, the correction surface based on the height data and the correction back surface height data, it is preferable to calculate the thickness data.

  According to this configuration, the standard surface height data and the standard back surface height data are calculated in advance by measuring the flat sample and the thickness sample. When measuring the solar cell wafer, the standard surface height data and the standard back surface height data are used to correct the front and back surface height data of the solar cell wafer, and the corrected front and back surface heights are corrected. The thickness data is calculated using the corrected front surface height data and the corrected back surface height data. As a result, the thickness data of the solar cell wafer can be obtained with high accuracy.

  (3) The standard surface height data is calculated by subtracting the surface height data of the flat sample from the surface height data of the thickness sample, and the standard back surface height data is It is preferable to calculate by subtracting the height data of the back surface from the height data of the back surface of the thickness sample.

  According to this configuration, it is possible to obtain standard surface height data in which the deviation of the height data associated with the inclination of the transport unit and the laser curvature is removed from the surface height data of the thickness sample. Further, standard back surface height data from which the deviation of the height data associated with the inclination of the transport unit and the laser curvature is removed from the back surface height data of the thickness sample can be obtained.

  (4) The transport unit uses the solar cell wafer, the flat sample, and the thickness sample as measurement samples, respectively, and transports the measurement sample at a constant transport speed. Each time it is transported, the measurement sample is imaged, and the height data calculation unit calculates the inclination of the front and back surfaces of the measurement sample from changes in the height data of the front and back surfaces in the transport direction of the measurement sample, It is preferable to correct the height data of the front and back surfaces of the measurement sample based on the calculated inclinations of the front and back surfaces so that the measurement points with respect to the transport direction are at regular intervals.

  According to this configuration, the height data can be corrected so that the measurement points due to the inclination of the front surface and the back surface of the measurement sample have a constant interval.

  (5) The parameter calculation unit sets an approximate back plane of the standard back surface height data as z2 (= a2x + b2y + c2), calculates coefficients a2, b2, c2 using the standard back surface height data, and calculates the calculated coefficients a2, b2, and c2 are calculated as back surface tilt correction parameters, a certain plane is set as z1 (= a1x + b1y + c1), and (the standard surface height data−z1) − (the standard back surface height data−z2) = the thickness sample. It is preferable that the coefficients a1, b1, and c1 are obtained so as to obtain the thickness data, and the obtained coefficients a1, b1, and c1 are calculated as the surface inclination correction parameters.

  According to this configuration, a flat surface showing the general inclination of the height data of the thickness sample when the surface of the flat sample is used as a reference surface, and the height data of the thickness sample when the back surface of the flat sample is used as a reference surface Are connected with a plane of the known thickness data of the thickness sample.

  (6) The height data correction unit subtracts the height data of the surface of the flat sample from the height data of the surface of the solar cell wafer, and further subtracts the plane z1 from the subtracted value, thereby correcting the corrected surface height. The height data of the back surface of the solar cell wafer is subtracted from the height data of the back surface of the flat sample, and the approximate back plane z2 is further subtracted from the subtracted value to obtain the corrected back surface height data. It is preferable.

  According to this configuration, the height data of the front and back surfaces of the flat sample is subtracted from the height data of the front and back surfaces of the solar cell wafer. It is possible to eliminate the height data shift due to the inclination and the height data shift due to the curvature of the light cutting line.

  Then, the height data of the front and back surfaces of the solar cell wafer from which the height data of the front and back surfaces of the flat sample are subtracted is further subtracted from the plane z1 and the approximate back plane z2.

  Therefore, the height data of the surface of the solar cell wafer with respect to the plane z1 = the height data of the back surface of the solar cell wafer with reference to the approximate back plane + thickness data. Therefore, the thickness of the solar cell wafer can be accurately obtained.

  (7) The measurement data acquisition unit includes a surface irradiation unit that irradiates a light cutting line on the surface of the solar cell wafer conveyed at a constant speed, and a light cutting line on the back surface of the solar cell wafer conveyed at a constant speed. , A front surface imaging unit that continuously images the surface of the solar cell wafer irradiated with the light cutting line by the front surface irradiation unit from above, and a light cutting line by the back surface irradiation unit. It is preferable to include a back surface imaging unit that continuously images the back surface of the solar cell wafer irradiated with the light from below at a constant period.

  According to this configuration, the front surface irradiation unit and the back surface irradiation unit are provided above and below the solar cell wafer, and the front surface imaging unit and the back surface imaging unit are provided above and below. The back side can be measured simultaneously.

  According to the present invention, since the thickness of the solar cell wafer is measured using the height data of the front and back surfaces of the solar cell wafer, not only the shape of the front and back surfaces of the solar cell wafer but also the thickness of the solar cell wafer. Can also be obtained. Therefore, the quality of the solar cell wafer can be determined using the thickness in addition to the saw mark. Moreover, since the surface measurement data and the back surface measurement data are obtained by continuously imaging the front surface and the back surface of the solar cell wafer at the same time, the height data and the thickness data of the solar cell wafer can be obtained at high speed.

(A) is a whole block diagram of the shape measuring apparatus by embodiment of this invention. (B) is a modification of (A). It is a block diagram of the shape measuring apparatus by embodiment of this invention. It is a flowchart which shows the process at the time of the shape measuring apparatus by embodiment of this invention measuring a standard sample. It is the flowchart which showed the process at the time of the shape measuring apparatus by embodiment of this invention measuring a wafer. (A), (B) is explanatory drawing of the calculation process of a back surface inclination correction parameter and a surface inclination correction parameter. (A) is a schematic diagram of the side surface of the shape measuring apparatus according to the embodiment of the present invention, and (B) is an enlarged view of the vicinity of the measurement stage of (A). It is the graph which showed the measurement result of the height data of the surface or back surface of a measurement sample, the left vertical axis shows the height data (μm) obtained by the measurement, and the right vertical axis shows the mounting surface of the measurement stage The angle of the measurement sample with respect to is shown. It is the figure which showed the shift | offset | difference of the measurement point of the optical cutting line in case the measurement sample with constant thickness inclines. It is the graph which showed the relationship between angle (phi) and thickness t after correction | amendment with respect to the measured value of t '= 200 micrometers. It is the figure which showed the shift | offset | difference of the measurement point of the optical cutting line in case the measurement sample whose thickness is not constant inclines. (A) shows a graph G11 showing the relationship of the actual thickness K to the angle φ1 in the case of φ2 = φ1 + 0.005 degrees, and a graph G12 showing the relationship of the actual thickness K−the measured thickness K ′ with respect to the angle φ1. Yes. (B) represents a graph G13 showing the relationship of the actual thickness K with respect to the angle φ1 when φ2 = φ1 + 0.1 degrees, and a graph G14 showing the relationship of the actual thickness K−the measured thickness K ′ with respect to the angle φ1. Yes. It is explanatory drawing of the measurement point shift | offset | difference correction | amendment performed by a height data calculation part. It is explanatory drawing of the measurement point shift | offset | difference correction | amendment performed by a height data calculation part. It is a flowchart which shows the detail of the calculation process of height data. It is a flowchart which shows the detail of a search process. It is a flowchart which shows the detail of a gravity center calculation process. It is the timing chart which showed the process of the flowchart of FIG. 14 in time series. It is the figure which showed notionally the flow of the process of a search process and a gravity center calculation process. It is the figure which showed an example of the light section line image. It is the graph which showed distribution of the luminance value when centering on the maximum luminance pixel in the i-th line. It is explanatory drawing of the process which calculates height data. It is the figure which showed typically the installation state of a light source and a camera.

  Hereinafter, a shape measuring apparatus according to an embodiment of the present invention will be described. FIG. 1A is an overall configuration diagram of a shape measuring apparatus according to an embodiment of the present invention. As shown in FIG. 1A, the shape measuring apparatus includes a light source 121 (an example of a front surface irradiation unit), a light source 131 (an example of a back surface irradiation unit), a camera 122 (an example of a front surface imaging unit), and a camera 132 (a back surface). An example of an imaging unit), a conveyance belt 3 and the like are provided.

  In FIG. 1A, the y direction indicates the conveyance direction of the measurement sample 500 by the conveyance belt 3. The x direction indicates a direction orthogonal to the y direction and parallel to the horizontal plane. The z direction indicates a height direction orthogonal to each of the x direction and the y direction. In the present embodiment, a standard sample 510 and a solar cell wafer (hereinafter referred to as “wafer”) 520 exist as the measurement sample 500. As the standard sample 510, there are a flat sample 511 whose front and back surfaces are flat and a thickness sample 512 whose thickness is known.

  That is, this shape measuring apparatus measures the flat sample 511 and the thickness sample 512, and calculates | requires the surface inclination correction parameter and back surface inclination correction parameter mentioned later previously. Then, the wafer 520 to be measured is measured, the height data of the front surface and the back surface of the wafer 520 is obtained, the obtained height data is corrected using the front surface tilt correction parameter and the back surface tilt correction parameter, and each of the wafers 520 is corrected. Find position thickness data.

  As the flat sample 511 and the thickness sample 512, those having the same surface area as the surface area of the wafer 520 to be measured are employed, for example, those having a surface area of 150 mm square. Further, as the thickness sample 512, a sample having a thickness comparable to that of the wafer 520 to be measured is employed, for example, a sample having a thickness of 200 μm. Note that if a flat sample 511 having a thickness of 200 μm is prepared, it is not necessary to prepare the flat sample 511 and the thickness sample 512 separately.

  However, since a warp or the like occurs when the thickness is reduced to about 200 μm, it is not easy to create a flat sample 511 having the same thickness as the wafer 520, and the cost increases. Therefore, in this embodiment, a flat sample 511 having a thickness larger than that of the wafer 520 but having a flat front and back surfaces and a thickness sample 512 having the same level as the wafer 520 are used.

  The light source 121 irradiates light on the surface of the measurement sample 500 so that the camera 122 spreads in a fan shape toward a region where the camera 122 captures a light section line image. That is, the light source 121 irradiates the surface of the measurement sample 500 conveyed at a constant speed with the light cutting line CL. The light source 131 irradiates light on the back surface of the measurement sample 500 conveyed at a constant speed so as to spread in a fan shape toward the area where the camera 132 captures the light section line image. That is, the light source 131 irradiates the light cutting line CL to the back surface of the measurement sample 500.

  In the present embodiment, the light cutting line CL is formed, for example, such that the coordinate in the y direction is located at a predetermined position y1, and the longitudinal direction is substantially parallel to the x direction. Thus, since the optical cutting line CL is irradiated to the entire width direction (x direction) of the front and back surfaces of the measurement sample 500 by the optical cutting line CL, the measurement sample 500 is moved at a constant speed in the y direction by the conveyor belt 3. By transporting and continuously capturing the optical cutting line images with the cameras 122 and 132, the height data of the entire front and back surfaces of the measurement sample 500 can be obtained at high speed.

  Each of the light sources 121 and 131 includes a cylindrical housing, and for example, a semiconductor laser and an optical system are provided inside the housing. The optical system is provided on the emission side of the semiconductor laser, and spreads and emits the laser beam emitted from the semiconductor laser in a fan shape.

  The support member 41 is arranged on the upstream side of the measurement stage 600 and supports the light source 121 so that the light source 121 is positioned at the center of the measurement stage 600 in the width direction (x direction). The support member 41 irradiates the surface of the measurement stage 600 with light from the light source 121 obliquely upward so that the elevation angle θ of the light with respect to the optical axis 701 when viewed from the x direction is a constant angle. The light source 121 is supported.

  The measurement stage 600 is a rectangular region having the same size as the measurement sample 500, and is positioned directly below the measurement sample 500, as shown in FIG.

  The support member 43 is disposed on the downstream side of the measurement stage 600 and supports the light source 131 so that the light source 131 is positioned at the center in the width direction (x direction) of the measurement stage 600. The support member 43 then irradiates the back surface of the measurement stage 600 with light from the light source 131 obliquely downward so that the elevation angle θ of the light with respect to the optical axis 702 when viewed from the x direction is a constant angle. The light source 131 is supported. That is, the elevation angles of the light sources 121 and 131 are made equal.

  The camera 122 is arranged on the upper side of the measurement stage 600 so that the optical axis 701 is located at the center of gravity of the measurement stage 600 and is parallel to the z direction. Then, the camera 122 continuously captures the surface of the measurement sample 500 irradiated with the light cutting line CL from above at a constant period, and acquires a light cutting line image.

  The camera 132 is disposed on the lower side of the measurement stage 600 so that the optical axis 702 is located at the center of gravity of the measurement stage 600 and is parallel to the z direction. Then, the camera 132 continuously captures the back surface of the measurement sample 500 irradiated with the light cutting line CL from below at a constant period, and acquires a light cutting line image.

  In the present embodiment, the cameras 122 and 132 have the same angle of view in the x direction and the y direction, respectively. In addition, the cameras 122 and 132 are disposed at positions where the distance from the measurement stage 600 is substantially equal. Therefore, the optical section line images captured by the cameras 122 and 132 are rectangular images of the same size.

  Here, the cameras 122 and 132 are configured by a CMOS camera capable of capturing an image at a predetermined frame rate (for example, 250 fps), and convert analog image data of the captured optical section line image into digital image data. To do.

  The conveyor belts 3 are endless belts whose longitudinal direction is the y direction, and a pair of the conveyor belts 3 are arranged with a constant interval in the x direction. The conveyor belt 3 conveys the measurement sample 500 in the y direction at a constant speed. In this way, the front and back surfaces of the measurement sample 500 can be simultaneously measured by providing the pair of transport belts 3 and 3 at a constant interval in the x direction.

  The support member 42 is erected at the center of the measurement stage 600 in the y direction, and supports the cameras 122 and 132.

  In FIG. 1A, the light source 121 is disposed on the upstream side of the measurement stage 600 and the light source 131 is disposed on the downstream side of the measurement stage 600. However, the present invention is not limited to this. That is, as illustrated in FIG. 1B, the light source 131 may be disposed on the upstream side of the measurement stage 600 and the light source 121 may be disposed on the downstream side of the measurement stage 600.

  FIG. 2 is a block diagram of the shape measuring apparatus according to the embodiment of the present invention. As shown in FIG. 2, the shape measuring apparatus includes a conveyance unit 110, measurement data acquisition units 120 and 130, a control unit 140, a parameter storage unit 150, and an operation unit 160.

  The conveyance unit 110 includes a conveyance roller that stretches the conveyance belts 3 and 3 and the conveyance belts 3 and 3 illustrated in FIG. 1, a motor that drives the conveyance rollers, and the like. In the transport unit 110, the transport roller is rotated at a constant angular speed and the transport belts 3 and 3 are rotated at a constant speed under the control of the transport control unit 141.

  The measurement data acquisition unit 120 includes a light source 121 and a camera 122 shown in FIG. 1A and a measurement data calculation unit 123 built in the camera 122. Then, the measurement data acquisition unit 120 irradiates the surface of the measurement sample 500 with the light cutting line CL from an obliquely upward direction, and acquires surface measurement data of the surface of the measurement sample 500.

  The measurement data acquisition unit 130 includes a light source 131 and a camera 132 shown in FIG. 1A and a measurement data calculation unit 133 built in the camera 132. Then, the measurement data acquisition unit 130 irradiates the optical cutting line CL obliquely downward with respect to the back surface of the measurement sample 500 to acquire surface measurement data on the surface of the measurement sample 500.

  The measurement data calculation unit 123 calculates surface measurement data by executing a search process and a centroid calculation process for each light section line image captured by the camera 122. Here, the search process is a process of setting a plurality of lines in parallel with the transport direction in the light section line image and searching for the maximum luminance pixel of each line. The center-of-gravity calculation process is a process of calculating the center-of-gravity coordinates of the maximum luminance pixel of each line in units of sub-pixels based on the maximum luminance pixel of each line searched by the search process. And the measurement data calculation part 123 calculates the gravity center coordinate obtained by the gravity center calculation process as surface measurement data.

  Similar to the measurement data calculation unit 123, the measurement data calculation unit 133 performs search processing and centroid calculation processing on each light section line image captured by the camera 132 to calculate back surface measurement data.

  The control unit 140 is configured by a computer, for example, and controls the entire shape measuring apparatus. Specifically, the control unit 140 includes a conveyance control unit 141, a measurement control unit 142, a height data calculation unit 143, a parameter calculation unit 144, a height data correction unit 145, and a thickness data calculation unit 146.

  The conveyance control unit 141 controls the conveyance unit 110 so that the conveyance belt 3 is driven at a constant conveyance speed. Here, as the conveyance speed, if the width of the light cutting line CL in the Y direction is α, the period of the cameras 122 to 132 is 1/250 = 0.004 s, and when set to α / 0.004, the measurement sample Since 500 can be scanned without a gap, for example, α / 0.004 may be set. It is assumed that wafer 520 is placed on transfer unit 110 so that the saw mark is in the y direction. Therefore, the wafer 520 is transported along the saw mark direction, and the light cutting line CL is irradiated in a direction substantially orthogonal to the saw mark direction.

  The measurement control unit 142 controls the measurement data acquisition units 120 and 130. Specifically, the measurement control unit 142 controls the cameras 122 and 132 so that the cameras 122 and 132 capture the light section line image at a constant frame rate (for example, 250 fps). In addition, the measurement control unit 142 controls the light sources 121 and 131 so that the light sources 121 and 131 emit a predetermined amount of light.

  Based on the surface measurement data calculated by the measurement data calculation unit 123 and the elevation angle θ of the light source 121, the height data calculation unit 143 of the measurement sample 500 for one column corresponding to the irradiation position of the light cutting line CL. By repeatedly executing the process of calculating the height data, the height data of each position on the surface of the measurement sample 500 is calculated. Further, the height data calculation unit 143 performs the same process as the front surface measurement data on the back surface measurement data calculated by the measurement data calculation unit 133, and obtains height data of each position on the back surface of the measurement sample 500. calculate.

  Further, the height data calculation unit 143 calculates the inclinations of the front and back surfaces of the measurement sample 500 from changes in the height data of the front and back surfaces in the transport direction of the measurement sample 500, and based on the calculated inclinations of the front and back surfaces. Then, the measurement point deviation correction is performed to correct the height data of the front and back surfaces of the measurement sample 500 so that the measurement points with respect to the transport direction are at regular intervals.

  12 and 13 are explanatory diagrams of the measurement point deviation correction performed by the height data calculation unit 143. FIG. In the present embodiment, obtaining the height of the measurement sample 500 at the measurement points MP (0), MP (1), MP (2),... Arranged at a constant interval d in the y direction. It is aimed. However, when the surface of the measurement sample 500 is inclined with respect to the measurement stage 600, the irradiation position of the light to the measurement sample 500 is measured at the measurement points MP (0), MP (1), MP (2),. Does not match and shifts. In the example of FIG. 12, at time t = 0, the measurement sample 500 is irradiated with light at the measurement point MP (0), and the height data z0 is calculated. Next, at time t = 1 when the measurement sample 500 is conveyed by the interval d, it is desired to obtain the height data z1 of the measurement point MP (1). Since (1), the measurement point is shifted from MP (1) to MP ′ (1).

  Furthermore, at time t = 2 when the measurement sample 500 is conveyed by the interval d, it is desired to obtain the height data z2 of the measurement point MP (2). However, the irradiation position of the light is measured at the measurement point MP ′ ( 2), the measurement point is shifted from MP (2) to MP ′ (2).

  Therefore, the height data calculation unit 143 uses the measurement point MP ′ (1) and the next measurement point as shown in FIG. 13 in order to correct the deviation of the measurement points MP (0) to MP (1). The height data z1 of the measurement point MP (1) is obtained by obtaining an inclination with a certain measurement point MP ′ (2) and performing linear interpolation using the inclination.

  Specifically, the height data calculation unit 143 obtains the height data z1 of the measurement point MP (1) using Expression (1).

z1 = ((z2′−z1 ′) / d2) · (d−d1) + z1 ′ (1)
The height data calculation unit 143 corrects the height data for the other measurement points MP using Equation (1).

  Thereby, even if the measurement sample 500 is inclined, the height data z0, z1 of the measurement points MP (0), MP (1), MP (2),... Arranged in the y direction at a constant interval d. , Z2,... As a result, the measurement point shift due to the inclination of the measurement sample 500 can be removed.

  Returning to FIG. 2, the parameter calculation unit 144 subtracts the surface height data of the flat sample 511 from the surface height data of the thickness sample 512 calculated by the height data calculation unit 143 to calculate the standard surface height data. To do.

  In addition, the parameter calculation unit 144 subtracts the height data of the back surface of the flat sample 511 from the height data of the back surface of the thickness sample 512 calculated by the height data calculation unit 143 to calculate the standard back surface height data.

  Then, the parameter calculation unit 144 sets z2 (= a2x + b2y + c2) as the approximate back plane of the standard back surface height data, and calculates the coefficients a2, b2, and c2 as the back surface inclination correction parameters. Further, the parameter calculation unit 144 sets a certain plane as z1 (= a1x + b1y + c1), and the coefficients a1, b1 so that the standard table plane height data−z1 = standard back surface height data−z2 = thickness sample 512 thickness data. , C1 and the obtained coefficients a1, b1, c1 are calculated as surface inclination correction parameters. The back surface tilt correction parameter and the front surface tilt correction parameter are stored in the parameter storage unit 150.

  The height data correction unit 145 corrects the front surface height data and the back surface height data of the wafer 520 calculated by the height data calculation unit 143 using the front surface tilt correction parameter and the back surface tilt correction parameter. Height data and corrected back surface height data are calculated.

  Specifically, the height data correction unit 145 subtracts the surface height data of the flat sample 511 from the surface height data of the wafer 520, and further subtracts the plane z1 from the subtracted value, thereby correcting the surface height data. Ask for.

  Further, the height data correction unit 145 subtracts the height data of the back surface of the flat sample 511 from the height data of the back surface of the wafer 520, and further subtracts the approximate back plane z2 from the subtracted value, thereby correcting the back surface height data. Ask for.

  The thickness data calculation unit 146 calculates thickness data at each position of the wafer 520 based on the corrected front surface height data and the corrected back surface height data.

  The operation unit 160 includes various operation buttons and a display panel, receives operation inputs from the operator, and displays measurement results of the measurement sample 500 and the like.

  Next, problems of the shape measuring apparatus according to the embodiment of the present invention will be described. FIG. 6A is a schematic side view of the shape measuring apparatus according to the embodiment of the present invention, and FIG. 6B is an enlarged view of the vicinity of the measurement stage 600 in FIG. FIG. 7 is a graph showing the measurement result of the height data of the front or back surface of the measurement sample 500. The left vertical axis shows the height data (μm) obtained by the measurement, and the right vertical axis shows the measurement. The angle of the measurement sample 500 with respect to the mounting surface 601 of the stage 600 is shown. The vertical axis indicates the moving distance (mm) of the measurement sample 500, and the movement distance = 0 indicates the position of the measurement sample 500 in the y direction when the measurement is started. A graph 71 indicates an angle, and a graph 72 indicates height data.

  As shown in FIG. 6B, the conveyance belt 3 gradually increases in height near the measurement stage 600 toward the measurement stage 600, becomes flat on the measurement stage 600, and increases as the distance from the measurement stage 600 increases. The trapezoidal shape is such that the thickness gradually decreases.

  Therefore, the measurement sample 500 is transported while being inclined obliquely upward and to the right with respect to the mounting surface 601, then transported in parallel with the placement surface 601, and then transported while being inclined obliquely downward and to the right. Is done. The height of the optical cutting line CL irradiated to the measurement sample 500 with respect to the measurement stage 600 gradually decreases while the measurement sample 500 is conveyed obliquely upward to the right with respect to the mounting surface 601, and the measurement sample 500 Is substantially constant while being transported in parallel to the mounting surface 601, and gradually increases while the measurement sample 500 is transported while being tilted to the lower right with respect to the mounting surface 601.

  Therefore, as shown in the graph 72 of FIG. 7, the height data of the measurement sample 500 changes while drawing a downward convex bathtub-like curve as the movement distance increases. Note that the minute fluctuations in the height data of the graph 72 are caused by the backlash of the conveyor belt 3 and the like. As described above, when the transport unit 110 is inclined, the height data of the measurement sample 500 cannot be accurately measured.

  FIG. 8 is a diagram showing the deviation of the measurement point MP when the measurement sample 500 having a constant thickness is inclined. In FIG. 8, the elevation angle θ indicates the angle formed between the direction of light from the light sources 121 and 131 and the optical axes 701 and 702 of the cameras 122 and 132. An angle φ indicates an inclination angle of the measurement sample 500 with respect to the placement surface 601.

  In this case, since the measurement sample 500 is inclined by the angle φ with respect to the placement surface 601, the measurement point MP is shifted to the measurement point MP ′ that is the upstream position in the y direction. As a result, the thickness of the measurement sample 500 is shifted to t ′ instead of t which is the original thickness. In FIG. 8, t ′ is represented by the equation (2).

t´ = ((1 / cosφ-1) / (tanθ ・ tanφ + 1) +1) ・ t (2)
Therefore, t, which is the original thickness of the measurement sample 500, is expressed by Equation (3).

t = ((sinφ ・ tanθ + cosφ) / (sinφ ・ tanθ + 1)) ・ t (3)
From the graph 71 shown in FIG. 7, the inclination of the measurement sample 500 with respect to the placement surface 601 is in the range of −0.5 degrees to +0.5 degrees. The value (t)) is within −0.01 μm to +0.01 μm.

  FIG. 9 is a graph showing the relationship between the angle φ and the corrected thickness t with respect to the measurement value of t ′ = 200 μm. In FIG. 9, the graph which shows the relationship between (phi) and t about each of (theta) = 80,82,84,86,88 degree is drawn. It can be seen that the thickness error is within 0.01 μm within the range of φ = −0.5 degrees to +0.5 degrees.

  In FIG. 8, the case where the thickness of the measurement sample 500 is constant has been described. However, the thickness of the wafer 520 to be measured is not constant in the transport direction and has some variation. FIG. 10 is a diagram showing the deviation of the measurement point MP of the light section line CL when the measurement sample 500 having a non-uniform thickness is inclined.

  In FIG. 10, the angle of the front surface of the measurement sample 500 with respect to the mounting surface 601 is φ1, and the angle of the back surface with respect to the mounting surface 601 is φ2. In this case, the measurement point MP (1) on the front surface is shifted to MP ′ (1), and the measurement point MP (2) on the back surface is shifted to MP ′ (2). Therefore, the measured value h1 ′ of the surface height data is represented by the formula (4). Further, the measured value h2 ′ of the height data on the back surface is expressed by the equation (5).

h1´ = (d ・ tanφ1 / (tanθ ・ tanφ1 + 1)) + h (4)
h2´ = d ・ tanφ2 / (tanθ ・ tanφ2 + 1) (5)
Therefore, the measured thickness K ′ is obtained by h1′−h2 ′.

  As described above, when the inclinations of the front and back surfaces of the measurement sample 500 are different from each other, the actual thickness K in the normal direction at the measurement point MP (2) is simply obtained by simply obtaining the measurement thickness K ′ by h1′−h2 ′. Cannot be asked.

  This actual thickness K is expressed by equation (6).

| h + d ・ (tanφ2-tanφ1) / (tanφ1 ・ tanφ2 + 1) | (tan ² φ2 + 1) ^1/2 (6)
Here, φ1 and φ2 are obtained from the equations (4) and (5), d is an interval between measurement points and is known, and if h is also known, an actual thickness K is obtained.

  FIG. 11A is a graph G11 showing the relationship of the actual thickness K to the angle φ1 when φ2 = φ1 + 0.005 degrees, and a graph G12 showing the relationship of the actual thickness K−the measured thickness K ′ with respect to the angle φ1. Represents.

  FIG. 11B shows a graph G13 showing the relationship of the actual thickness K with respect to the angle φ1 when φ2 = φ1 + 0.1 degrees, and a graph G14 showing the relationship of the actual thickness K−the measured thickness K ′ with respect to the angle φ1. Represents.

  11A and 11B, the left vertical axis indicates the actual thickness K, the right vertical axis indicates the actual thickness K−the measured thickness K ′, and the horizontal axis indicates φ1. As shown in the graphs G11 and G13, it can be seen that the actual thickness K decreases as φ1 moves away from zero. Further, as shown in the graphs G12 and G14, it can be seen that the actual thickness K−the measured thickness K ′ decreases as φ1 increases from −0.5 degrees to 0.5 degrees.

  As shown in FIGS. 11A and 11B, when the difference between φ1 and φ2 is 0.05 degrees and 0.1 degrees, φ1 has an inclination within the range of −0.5 to +0.5. It can be seen that the variation in thickness is within 0.01 μm.

  In this embodiment, the height data shift (hereinafter referred to as shift (1)) due to the inclination of the conveyor belt 3 shown in FIG. 6B and the measurement point shift shown in FIGS. It is an object of the present invention to correct the resulting height data shift (hereinafter referred to as shift (2)). In addition, the light cutting line CL may be bent and irradiated in an arc shape due to the characteristics of the optical system included in the light sources 121 and 131. In this case, the height data of the measurement sample 500 deviates from the original height data. In the present embodiment, it is an object to correct height data shift (hereinafter referred to as shift (3)) caused by the curvature of the light cutting line CL.

  Next, the operation of the shape measuring apparatus according to this embodiment will be described. FIG. 3 is a flowchart showing processing when the shape measuring apparatus according to the embodiment of the present invention measures the standard sample 510. First, the flat sample 511 is conveyed toward the measurement stage 600, the optical cutting line CL is irradiated to the flat sample 511, and the optical cutting line image is captured by the cameras 122 and 132 each time the flat sample 511 is conveyed by the distance d. The

  Then, the measurement data calculation units 123 and 133 obtain the surface measurement data and the back surface measurement data of the flat sample 511 from each light section line image, and the height data calculation unit 143 calculates the flat sample 511 from the surface measurement data and the back surface measurement data. The height data of each position of the front surface and the back surface is calculated (S1).

  Next, the height data calculation unit 143 performs measurement point deviation correction on the height data at each position on the front and back surfaces of the flat sample 511 obtained in S1 (S2).

  Thereby, deviation (2) can be corrected. Thereby, the height data of each measurement point arranged in a predetermined row × predetermined column is obtained.

  Next, each time the thickness sample 512 is conveyed toward the measurement stage 600, the optical cutting line CL is irradiated to the thickness sample 512, and the thickness sample 512 is conveyed by the distance d, the optical cutting line image is captured by the cameras 122 and 132. Is done. Then, the measurement data calculation units 123 and 133 obtain the surface measurement data and the back surface measurement data of the thickness sample 512 from each light cutting line image, and the height data calculation unit 143 determines the positions of the front and back surfaces of the thickness sample 512. Height data is calculated (S3).

  Next, the height data calculation unit performs measurement point deviation correction on the height data of the positions of the front and back surfaces of the thickness sample obtained in S3 (S4). As a result, the deviation (2) is corrected.

  Next, the parameter calculation unit 144 calculates each surface of the flat sample 511 subjected to the measurement point deviation correction in S2 from the height data of each position of the surface of the thickness sample 512 subjected to the measurement point deviation correction in S4. The position height data is subtracted to calculate the standard surface height data (S5). Thereby, the height data of the surface of the thickness sample 512 when the surface of the flat sample 511 is used as a reference plane is obtained. Accordingly, the deviation (1) and the deviation (3) are removed from the height data of the thickness sample 512. That is, both the height data of the surface of the thickness sample 512 and the height data of the surface of the flat sample 511 are affected by the shift (1) and the shift (3). Therefore, the surface height data of the thickness sample 512−the surface height data of the flat sample 511 cancels the influence of the deviation (1) and the deviation (3), and the deviation (1) and the deviation (3) are removed. The

  Note that subtracting the height data of each position of the flat sample 511 from the height data of each position of the thickness sample 512 means subtracting the height data of the measurement point of the same coordinate in the thickness sample 512 and the flat sample 511. means. In the following, it is assumed that the processing applied to height data and thickness data is performed on height data and thickness data having the same coordinates.

  Next, the parameter calculation unit 144 uses the height data at each position on the back surface of the thickness sample 512 on which the measurement point deviation correction has been performed in S4, to calculate each of the back surface of the flat sample 511 on which the measurement point deviation correction has been performed in S2. The position height data is subtracted to calculate standard back surface height data (S6). Thereby, the height data of the back surface of the thickness sample 512 when the back surface of the flat sample 511 is used as a reference surface is obtained, and the shift (1) and the shift (3) are corrected.

  Next, the parameter calculation unit 144 calculates the approximate back plane z2 of the standard back surface height data. FIGS. 5A and 5B are explanatory diagrams of calculation processing of the back surface inclination correction parameter and the front surface inclination correction parameter.

  It is assumed that the standard surface height data obtained in S5 is distributed as a plane A1 in FIG. Further, it is assumed that the standard back surface height data obtained in S6 is distributed as a surface A2 in FIG. Therefore, the approximate back plane z2 is represented as a plane A2.

  As described above, the inclinations of the surface A1 and the surface A2 are different because the global inclination of the surface of the thickness sample 512 when the surface of the flat sample 511 is used as a reference surface and the back surface of the flat sample 511 is a reference This occurs because the global inclination of the thickness sample 512 is shifted from the surface. This deviation in inclination occurs due to the way the thickness sample 512 is placed on the conveyor belt 3, the positions of the light sources 121 and 131, the positions of the cameras 122 and 132, and the like.

  Therefore, in the present embodiment, first, an approximate back plane z2 (= a2 · x + b2 · y + c2) is obtained from the standard back surface height data by the least square method as shown in Expression (7).

E = Σ (z2−a2 · x−b2 · yc2) ² (7)
Then, the coefficients a2, b2, and c2 that minimize E are obtained as back surface tilt correction parameters.

  Specifically, coefficients a2, b2, and c2 that satisfy ∂E / ∂a2 = 0, ∂E / ∂b2 = 0, and ∂E / ∂c2 = 0 are obtained.

  Next, the parameter calculation unit 144 obtains standard back surface height data−approximate back plane z2. Thereby, as shown in FIG. 5B, the approximate back plane z2 is set to the plane A4 on the xy plane, and the height data of the back surface of the thickness sample 512 when the plane A4 is taken as the reference plane is obtained. It is done.

  Next, the parameter calculation unit 144 sets a certain plane as a plane z1 (= a1 · x + b1 · y + c1), and (standard surface height data−z1) − (standard back surface height data−z2) = known thickness data Thus, the coefficients a1, b1, and c1 are obtained (S8). These coefficients a1, b1, and c1 are surface inclination correction parameters.

  Here, the known thickness data is thickness data assigned in advance to each position of the thickness sample 512. This thickness data may have the same value for each position, or may have different values.

  Thereby, the standard surface height data when the plane z1 is used as a reference plane is further added to the value obtained by adding known thickness data to the height data of the back surface of the thickness sample 512 when the plane A4 is used as the reference plane. A value is calculated as the height data of the surface of the thickness sample 512. As a result, as shown in FIG. 5 (B), a plane that is a reference plane for the height data of the surface of the thickness sample 512 at a position spaced about the known thickness data with respect to the approximate back plane z2 in the z-direction side. z1 can be set.

  In calculating the surface inclination correction data, since there are three variables a1, b1, and c1, it is sufficient to use at least three points of thickness data, standard back surface height data, and standard surface height data.

  Returning to FIG. 3, the parameter calculation unit 144 causes the parameter storage unit 150 to store the back surface inclination correction parameters (a2, b2, c2) obtained in S7 and the front surface inclination correction parameters (a1, b1, c1) obtained in S8. .

  Next, a process for measuring the wafer 520 will be described. FIG. 4 is a flowchart showing processing when the shape measuring apparatus according to the embodiment of the present invention measures the wafer 520.

  First, the wafer 520 is transported toward the measurement stage 600, the optical cutting line CL is irradiated onto the wafer 520, and the optical cutting line image is captured by the cameras 122 and 132 each time the wafer 520 is transported by the distance d. Then, the height data calculation unit 143 calculates the height data at each position on the front and back surfaces of the wafer 520 from each cutting line image (S21).

  Next, the height data calculation unit 143 subtracts the height data of the surface of the flat sample 511 obtained in S1 of FIG. 3 from the height data of the surface of the wafer 520 calculated in S21 (S22). Thereby, the height data of the surface of the wafer 520 when the surface of the flat sample 511 is used as a reference plane is obtained. Thereby, deviation (1) and deviation (3) are corrected. That is, both the height data of the surface of the wafer 520 and the height data of the surface of the flat sample 511 are affected by the shift (1) and the shift (3). Therefore, the influence of the deviation (1) and the deviation (3) is canceled by the height data of the surface of the wafer 520-the height data of the surface of the flat sample 511, and the deviation (1) and the deviation (3) are removed. .

  Next, the height data calculation unit 143 subtracts the height data of the back surface of the flat sample 511 from the height data of the back surface of the wafer 520 calculated in S21 (S23). Thereby, the height data of the back surface of the wafer 520 when the back surface of the flat sample 511 is used as a reference surface is obtained. As a result, the shift (1) and the shift (3) are removed as in S22.

  Next, the height data calculation unit 143 performs measurement point deviation correction on the front surface height data obtained in S22 and the back surface height data obtained in S23 (S24). Thereby, the shift (2) is removed.

  Next, the height data correction unit 145 calculates the corrected surface height data by subtracting the plane z1 from the surface height data on which the measurement point deviation correction has been performed in S24 (S25). Thereby, the height data of the wafer 520 when the plane z1 is used as a reference plane is obtained.

  Next, the height data correction unit 145 calculates the corrected back surface height data by subtracting the approximate back surface z2 from the back surface height data that has been subjected to the measurement point deviation correction in S24 (S26). Thereby, the height data of the wafer 520 when the approximate back plane z2 is used as a reference plane is obtained.

  Next, the height data correction unit 145 calculates the thickness data of the wafer 520 from the corrected front surface height data−corrected back surface height data (S27). The plane z1 has known thickness data as a negative thickness component. Therefore, the corrected surface height data obtained from the surface height data-plane z1 in S25 has a known thickness data value. Accordingly, by calculating the corrected front surface height data-corrected back surface height data in S27, the thickness data of the wafer 520 in consideration of the known thickness data can be obtained.

  Next, details of the height data calculation processing in S1 and S3 in FIG. 3 and S21 in FIG. 4 will be described. FIG. 14 is a flowchart showing details of the height data calculation process. In FIG. 14, it is assumed that the cameras 122 and 132 continuously capture X optical cutting line images. First, the transport unit 110 is driven by the transport control unit 141, and transport of the measurement sample 500 is started (S141).

  Next, the cameras 122 and 132 capture the image of the measurement sample 500 and acquire image data of the light section line image of the first frame (S142 (1)).

  FIG. 19 is a diagram illustrating an example of a light section line image. The optical section line image shown in FIG. 19 is image data having M pixels in the x direction (vertical direction) and N pixels in the y direction (transport direction), that is, M rows × N columns. And In addition, the pixel value of each pixel is represented by, for example, 256 gradations from 0 to 255. Hereinafter, the pixel value is described as a luminance value.

  As shown in FIG. 19, it can be seen that a linear light section line CL appears along the x direction. Returning to FIG. 14, in S <b> 142 (2), the cameras 122 and 132 capture the image of the measurement sample 500 and acquire image data of the optical cutting line image of the second frame. At the same time, the measurement data calculation units 123 and 133 execute search processing on the image data of the light section line image of the first frame.

  In S <b> 142 (3), the cameras 122 and 132 capture the measurement sample 500 and acquire image data of the light section line image of the third frame. At the same time, the measurement data calculation units 123 and 133 perform search processing on the image data of the light section line image of the second frame to search for the maximum luminance pixel of each line of the light section line image. At the same time, the measurement data calculation units 123 and 133 execute a centroid calculation process on the light section line image of the first frame, and calculate the centroid coordinates of the maximum luminance pixel of each line in subpixel units.

  When S142 (3) is completed, the barycentric coordinates in the first column corresponding to the light section line CL appearing in the first frame are obtained.

  Thereafter, the same processing as S142 (3) is repeatedly executed from S142 (4) to S142 (X). When S142 (X) is completed, the barycentric coordinates of the first column to the X-2 column are obtained.

  In S142 (X + 1), since the imaging by the cameras 122 and 132 has been completed, only the search process and the centroid process are executed, and the centroid coordinates of the X-1th column are obtained.

  In S142 (X + 2), since the search process is also ended, only the centroid calculation process is executed, and the centroid coordinates of the Xth column are obtained.

  Through the pipeline processing described above, the center-of-gravity coordinates for one column are obtained each time a period in which the cameras 122 and 132 capture one frame of the light section line image has elapsed.

  FIG. 15 is a flowchart showing details of the search process. First, the measurement data calculation units 123 and 133 set the light cutting line image of the frame immediately before the current frame currently captured by the cameras 122 and 132 as the light cutting line image to be processed (S151).

  Next, the measurement data calculation units 123 and 133 substitute 0 for a variable i for indicating the line number of each line set on the light section line image, and initialize i (S152). In this case, as shown in FIG. 19, it can be seen that one line is set along the y direction on the light section line image. Note that one line is represented by i = 0 representing the first row of the optical section line image shown in FIG. 19, and i = 1 representing the second line of the optical section line image shown in FIG. Corresponding to one row of the light section line image, the variable i is associated with each row of the light section line image.

  Next, the measurement data calculation units 123 and 133 search for the maximum luminance pixel that is the pixel having the maximum luminance in the i-th line (S153). In this case, the measurement data calculation units 123 and 133 search for the maximum brightness pixel by sequentially setting, for example, the leftmost pixel to the rightmost pixel as the target pixel in the i-th line illustrated in FIG. Specifically, first, the leftmost pixel is set as a target pixel, and the luminance value and coordinates are stored in a buffer (not shown). Here, as the coordinates, an integer value indicating the number of pixels from the leftmost pixel can be adopted.

  Next, the pixel on the right is set as the pixel of interest, and if the luminance value of the pixel of interest is greater than or equal to the luminance value stored in the buffer, the buffer is updated with the luminance value and coordinates of the pixel of interest. On the other hand, when the luminance value of the target pixel is less than the luminance value stored in the buffer, the luminance value and coordinates stored in the buffer are not updated. Such processing is repeated, and finally the coordinates stored in the buffer are determined as the coordinates yp of the maximum luminance pixel, and the maximum luminance pixel of the i-th line is searched.

  The obtained coordinates yp of the maximum luminance pixel of the i-th line are stored in a buffer (not shown) in association with the variable i.

  Next, when the process for obtaining the coordinate yp of the maximum luminance pixel is completed for all lines (YES in S154), the measurement data calculation units 123 and 133 return the process and the coordinate yp of the maximum luminance pixel for all lines. Is not completed (NO in S154), i is incremented by 1 (S155), and the process returns to S153.

  That is, the measurement data calculation units 123 and 133 obtain the coordinates yp of the maximum luminance pixels of all the lines of the light section line image shown in FIG. 19 by repeating the processes of S153 to S155.

  FIG. 16 is a flowchart showing details of the center of gravity calculation processing. First, the measurement data calculation units 123 and 133 set the light cutting line image of the frame immediately before the current frame currently captured by the cameras 122 and 132 as the light cutting line image to be processed (S161).

  Next, the measurement data calculation units 123 and 133 assign 0 to the variable i in the same manner as the search process, and initialize i (S162). Next, the measurement data calculation units 123 and 133 extract n peripheral pixels left and right around the maximum luminance pixel searched in the i-th line, and use the maximum luminance pixel and 2n peripheral pixels, The barycentric coordinate ysub of the maximum luminance pixel on the i-th line is obtained (S163).

  FIG. 20 is a graph showing the distribution of luminance values when the maximum luminance pixel in the i-th line is the center. In FIG. 20, the coordinate yp is the coordinate of the maximum luminance pixel, and the distribution of luminance values of a total of 17 pixels, 8 on the right and 8 on the left, is shown.

  Then, the measurement data calculation units 123 and 133 obtain the barycentric coordinates of the maximum luminance pixel using, for example, the following formula.

  Where ysub represents the barycentric coordinate of the maximum luminance pixel, yj represents the coordinate of the jth pixel in the i-th line, kj represents the luminance value of yj, yp represents the coordinate of the maximum luminance pixel, and kp is The luminance value of the maximum luminance pixel is shown, and n is an index for specifying the peripheral pixels.

  As a result, the barycentric coordinate ysub of the maximum luminance pixel in the i-th line is obtained with a value with a decimal point of one pixel or less, that is, in sub-pixel units. Note that the measurement data calculation units 123 and 133 determine in advance how many decimal places the centroid coordinate ysub of the maximum luminance pixel is to be obtained, and are rounded when the centroid coordinate ysub of the maximum luminance pixel exceeds the number of digits. , Truncation, or rounding up may be performed.

  In the example of FIG. 20, the y value of the peak of the graph indicated by the solid line is the barycentric coordinate ysub of the maximum luminance pixel. In the example of FIG. 20, the number of peripheral pixels is n = 8, but this is an example, and if it is 1 or more, another value may be appropriately adopted as long as the calculation amount does not become enormous. .

  Returning to FIG. 16, when the processing for obtaining the center-of-gravity coordinate ysub of the maximum luminance pixel is completed for all lines (YES in S164), the measurement data calculation units 123 and 133 return the processing to return the center-of-gravity coordinate ysub for all lines. Is not completed (NO in S164), i is incremented by 1 (S165), and the process returns to S163.

  In other words, the measurement data calculation units 123 and 133 obtain the barycentric coordinates ysub of the maximum luminance pixels of all the lines of the optical section line image shown in FIG. 20 by repeating the processes of S163 to S165.

  The barycentric coordinates ysub of the maximum luminance pixels of all lines are stored in a buffer (not shown) in association with the frame number of the light section line image and the variable i. Further, the center of gravity coordinate ysub of the surface of the measurement sample 500 obtained by the measurement data calculation unit 123 corresponds to the surface measurement data, and the center of gravity coordinate ysub of the measurement sample 500 obtained by the measurement data calculation unit 133 corresponds to the back surface measurement data.

  FIG. 17 is a timing chart showing the processing of the flowchart of FIG. 14 in time series. Periods T1 to T (X + 2) shown in FIG. 17 represent the time required for the cameras 122 and 132 to capture one frame of the optical section line image, that is, the frame period.

  In the period T1, S142 (1) shown in FIG. 14 is executed, and the light cutting line image of the first frame is captured. In the period T2, S142 (2) shown in FIG. 14 is executed, and the optical cutting line image of the second frame and the optical cutting line image search process of the first frame are simultaneously performed.

  In the period T3, S142 (3) shown in FIG. 14 is executed, and the optical cutting line image of the third frame, the optical cutting line image search process of the second frame, and the optical cutting line image of the first frame are processed. The center of gravity calculation process is simultaneously performed.

  Thereafter, until the period T (X), imaging of the optical cutting line image of the current frame, search processing of the optical cutting line image of the previous frame, and gravity center calculation processing of the optical cutting line image of the previous frame, Are performed simultaneously, and the barycentric coordinate ysub of the maximum luminance pixel for one column is calculated every time one period elapses.

  FIG. 18 is a diagram conceptually showing the flow of the search process and the center-of-gravity calculation process. A vertical axis shown in FIG. 18 indicates a time axis, and a scale is engraved for each frame period of the cameras 122 and 132. As described above, since the frame rates of the cameras 122 and 132 are 250 fps, the frame period is 4 msec. In FIG. 18, the number of lines set in the light cutting line image of one frame is 480 from i = 0 to 479.

  In the period T (n), the measurement sample 500 is imaged, and the optical section line image of the nth frame is acquired. Next, in the period T (n + 1), the measurement sample 500 being conveyed at a constant speed without stopping toward the left side is imaged, and the light section line image of the (n + 1) th frame is acquired. Next, in the period T (n + 2), the measurement sample 500 being transported at a constant speed without stopping toward the left side is imaged, and the light cutting line image of the (n + 2) th frame is acquired.

  In the period T (n), i = 0 to 479 lines are set for the optical section line image of the (n−1) th frame, and the maximum luminance pixel of each line is searched. I = 0 to 479 lines are set for the light section line image of the frame, and the barycentric coordinate ysub of the maximum luminance pixel of each line is calculated.

  In the periods T (n + 1) and T (n + 2), the same processing as that in the period T (n) is performed. Therefore, whenever the period T (n) elapses, the barycentric coordinate ysub of the maximum luminance pixel of i = 0 to 479 is obtained.

  Returning to FIG. 14, in S <b> 143, the height data calculation unit 143 calculates the center-of-gravity coordinates ysub of the maximum luminance pixels of one column calculated for the one-frame optical section line image calculated by the measurement data calculation units 123 and 133. Substituting each into the following formula, calculating height data, and arranging the calculated height data in the order of the variable i, the height data for one column is calculated.

h (μm) = R · ysub · cos θ / sin θ
Here, h represents height data, R represents visual field resolution, and θ represents the elevation angle of the light sources 121 and 131.

  FIG. 21 is an explanatory diagram of processing for calculating height data. FIG. 22 is a diagram schematically illustrating the installation state of the light source 121 and the camera 122. A square shown in FIG. 21 shows a light cutting line image, and a thick line shown in the square shows a light cutting line CL ′ drawn by the barycentric coordinates ysub of the maximum luminance pixels of all lines.

  As shown in FIG. 22, the elevation angle of the light source 121 is set to θ with reference to the z direction. Then, the position where the light emitted from the light source 121 enters the camera 122 moves back and forth in the y direction shown in FIG. 10 according to the height of the measurement sample 500. Therefore, the height data at each position on the surface of the measurement sample 500 can be obtained by substituting the barycentric coordinate ysub of the maximum luminance pixel into the above formula.

  In FIG. 21, the vertical length of the light section line image is ML (μm), the horizontal length is NL (μm), the vertical pixel number is M, and the horizontal pixel number is N. Then, the visual field resolution R is R = NL / N.

  Then, the height data calculation unit 143 arranges the M pieces of height data h obtained with respect to the barycentric coordinates ysub on the optical section line CL ′ in one row. Thereby, height data for one row of the measurement sample 500 is obtained.

  Returning to FIG. 14, in S <b> 144, the height data calculation unit 143 calculates the height data of the entire front and back surfaces of the measurement sample 500 by arranging the obtained height data for one column in the order of frame numbers. .

  In the example of FIG. 14, since the number of frames of the light section line image is X, height data in which X pieces of height data for one column are arranged is obtained. As described above, the height data of the front surface and the back surface of the measurement sample 500 is obtained. In this embodiment, the wafer 520 is a measurement target. The wafer 520 is manufactured by cutting a columnar ingot in a direction orthogonal to the longitudinal direction. Therefore, streaky irregularities called saw marks appear on the front and back surfaces of the wafer 520. The saw marks appear almost in parallel on the front surface and the back surface of the wafer 520. When judging whether the wafer 520 is good or bad, it is necessary to know how much the saw mark appears.

  Since the shape measuring apparatus obtains height data of the entire area of the front and back surfaces of the wafer 520 as described above, the saw mark appearing on the wafer 520 can be evaluated. Further, as shown in FIG. 14, the shape measuring apparatus performs imaging of the front and back surfaces of the wafer 520, search processing, and center-of-gravity calculation processing in a pipeline manner. Therefore, the surface measurement data and the back surface measurement data are obtained almost simultaneously with the completion of imaging of the entire front and back surfaces of the wafer 520, and the height data of the front and back surfaces of the wafer 520 can be calculated in real time.

  In addition, since the shape measuring apparatus measures the thickness of the wafer 520 using the height data of the front and back surfaces of the wafer 520, not only the shape of the front and back surfaces of the wafer 520 but also the thickness of the wafer 520 is obtained. be able to. Therefore, the quality of the wafer 520 can be determined using the thickness in addition to the saw mark.

  In the above description, the standard sample 510 is divided into the flat sample 511 and the thickness sample 512. However, when a flat sample having the same thickness as the wafer 520 is formed, this sample may be adopted as the standard sample 510.

  In this case, the height data of the front and back surfaces of the standard sample 510 represents the height data of the front and back surfaces of the flat sample 511 and the height data of the front and back surfaces of the thickness sample 512. The height data and the standard back surface height data are zero.

  Therefore, the height data correction unit 145 obtains corrected surface height data by subtracting the surface height data of the standard sample 510 from the surface height data of the wafer 520, and calculates the standard sample from the height data of the back surface of the wafer 520. The corrected back surface height data may be obtained by subtracting the back surface height data 510.

  In the above description, the thickness data is calculated using the standard sample 510. However, the present invention is not limited to this, and the height data of the front surface and the back surface of the wafer 520 is obtained. Thus, the thickness of the wafer 520 may be obtained.

3, 3 Conveying belt 110 Conveying unit 120, 130 Measurement data acquisition unit 121, 131 Light source 122, 132 Camera 123, 133 Measurement data calculating unit 140 Control unit 141 Conveying control unit 142 Measurement control unit 143 Height data calculating unit 144 Parameter calculation Unit 145 height data correction unit 150 parameter storage unit 160 operation unit 500 measurement sample 510 standard sample 511 flat sample 512 thickness sample 520 wafer 600 measurement stage 601 mounting surface

Claims (7)

Irradiate a light cutting line in a direction intersecting the transport direction of the solar cell wafer from an oblique direction with respect to the front and back surfaces of the solar cell wafer to be transported, continuously image the front and back surfaces of the solar cell wafer, and A measurement data acquisition unit for acquiring the front surface measurement data and the back surface measurement data of the battery wafer;
Based on the surface measurement data and the back surface measurement data acquired by the measurement data acquisition unit, the height data calculation unit that calculates the height data of each position on the surface of the solar cell wafer and the height data of each position on the back surface When,
A thickness data calculation unit that calculates thickness data at each position of the solar cell wafer based on the height data of the front surface and the height data of the back surface of the solar cell wafer calculated by the height data calculation unit. Shape measuring device. The height data calculation unit has a thickness having the same thickness as the solar cell wafer, and the height data of the front and back surfaces acquired by the measurement data acquisition unit measuring a flat sample with flat front and back surfaces. Based on the height data of the front and back surfaces obtained by measuring a known thickness sample, the standard surface height data and the standard back surface height data are calculated,
Based on the standard surface height data and the standard back surface height data, a parameter calculation unit that calculates a surface tilt correction parameter and a back surface tilt correction parameter;
The front surface height data and the back surface height data of the solar cell wafer calculated by the height data calculation unit are corrected using the front surface tilt correction parameter and the back surface tilt correction parameter, and the corrected front surface height data and A height data correction unit for calculating corrected back surface height data;
The shape measuring apparatus according to claim 1, wherein the thickness data calculation unit calculates the thickness data based on the corrected front surface height data and the corrected back surface height data. The standard surface height data is calculated by subtracting the surface height data of the flat sample from the surface height data of the thickness sample,
The shape measuring apparatus according to claim 2, wherein the standard back surface height data is calculated by subtracting the back surface height data of the flat sample from the back surface height data of the thickness sample. The transport unit uses the solar cell wafer, the flat sample, and the thickness sample as measurement samples, respectively, and transports the measurement sample at a constant transport speed.
The measurement unit images the measurement sample every time the measurement sample is transported a certain distance,
The height data calculation unit calculates the inclination of the front and back surfaces of the measurement sample from changes in the height data of the front and back surfaces in the conveyance direction of the measurement sample, and based on the calculated inclinations of the front and back surfaces The shape measuring apparatus according to claim 3, wherein the height data of the front surface and the back surface of the measurement sample is corrected so that the measurement points with respect to the direction are at regular intervals. The parameter calculation unit
An approximate back plane of the standard back surface height data is set as z2 (= a2x + b2y + c2), coefficients a2, b2, and c2 are calculated using the standard back surface height data, and the calculated coefficients a2, b2, and c2 are corrected for back surface inclination. As a parameter,
A plane is set as z1 (= a1x + b1y + c1),
(The standard surface height data-z1)-(The standard back surface height data-z2) = The coefficients a1, b1, c1 are determined so as to be the thickness data of the thickness sample, and the determined coefficients a1, b1, The shape measuring apparatus according to claim 3 or 4, wherein c1 is calculated as the surface inclination correction parameter. The height data correction unit
Subtracting the height data of the surface of the flat sample from the height data of the surface of the solar cell wafer, obtaining the corrected surface height data by further subtracting the plane z1 from the subtracted value,
6. The corrected back surface height data is obtained by subtracting the back surface height data of the flat sample from the back surface height data of the solar cell wafer, and further subtracting the approximate back surface z2 from the subtracted value. Shape measuring device. The measurement data acquisition unit
A surface irradiation unit for irradiating the surface of the solar cell wafer conveyed at a constant speed with a light cutting line;
A back side irradiation unit that irradiates a light cutting line to the back side of the solar cell wafer conveyed at a constant speed;
A surface imaging unit that continuously images the surface of the solar cell wafer irradiated with the light cutting line by the surface irradiation unit at a constant period from above;
The shape measuring apparatus according to claim 1, further comprising: a back surface imaging unit that continuously images the back surface of the solar cell wafer irradiated with the light cutting line by the back surface irradiation unit at a constant cycle from below.

Images