JP5277986B2 - Arithmetic device, arithmetic program, surface shape measuring device, and surface shape measuring method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To surely detect a location, where unwrap failure has occurred, to a phase image in a surface shape measurement by a phase shift method. <P>SOLUTION: The embodiment for disclosing a surface shape measurement method includes: a measurement procedure for obtaining a plurality of stripe images by imaging a to-be-measured object several times on which a stripe pattern having a sinusoidal intensity distribution is projected while a phase of the stripe pattern is changed, projecting a code pattern for identifying respective stripes included in the stripe pattern to the to-be-measured object instead of the stripe pattern, imaging the to-be-measured object in the state, and obtaining a code image; an unwrap procedure for calculating a phase distribution of the stripe pattern deformed in response to a surface shape of the to-be-measured object by fitting a plurality of the stripe images to formula of the phase shift method, and unwrapping the phase distribution based on the code image; and a determination procedure for obtaining a difference between an unwrap image as the phase distribution after the unwrap and a smoothed image of the unwrap image, and determining a pixel having a difference larger than a threshold value in the unwrap image as an improper pixel. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a surface shape measuring apparatus using a phase shift method and its peripheral technology.

  Conventionally, a phase shift method as shown in Patent Document 1 is known as a method for measuring the surface shape of an object to be measured. In the phase shift method, a fringe pattern having a sinusoidal intensity distribution is projected onto a measurement object, and the measurement object is imaged while changing the phase of the fringe pattern. A fringe phase distribution (phase image) is obtained by applying to a predetermined arithmetic expression, and the surface shape of the measurement object is restored by unwrapping (phase connection) the phase image.

  Unwrapping is a process of adding 2nπ (n is an integer, hereinafter referred to as “offset number”) to each pixel of the phase image. The offset number n of each pixel depends on where the pixel corresponds to the aforementioned stripe pattern. For example, the offset number n of the pixel corresponding to the mth stripe included in the stripe pattern is n = (m -1). Various methods for making the distribution of the offset number n on the phase image known have been proposed, and one of them uses the spatial code method.

  In this method, instead of the above-described stripe pattern, a code pattern (binary code pattern or gray code pattern) for identifying individual stripes of the stripe pattern is projected onto a measurement object, and a code image captured in that state is projected. First, the distribution of the offset number n on the phase image is calculated.

  However, since the code pattern projected onto the measurement object is affected by the aberration and noise of the projection optical system, for example, when a binary code pattern is adopted as the code pattern, a black pattern and a white pattern to be projected side by side are used. There is a possibility that the boundary portion (edge) with the pattern becomes dull and the SN ratio of the boundary portion is lowered. In this case, the calculation accuracy of the offset number n regarding the boundary portion is low, and there is a high possibility that unwrapping will fail. If unwrapping fails, there is a problem because spike-like noise occurs in the phase image.

  SUMMARY OF THE INVENTION An object of the present invention is to reliably detect an unwrapping failure point with respect to a phase image in surface shape measurement by the phase shift method.

  In one aspect of the arithmetic device illustrating the present invention, a measurement object on which a fringe pattern having a sinusoidal luminance distribution is projected is imaged a plurality of times while changing the phase of the fringe pattern to obtain a plurality of fringe images. A measurement device that obtains a code pattern for identifying individual stripes included in the stripe pattern instead of the stripe pattern and projects the code pattern onto the measurement object, and images the measurement object in that state to obtain a code image The phase distribution of the fringe pattern deformed according to the surface shape of the measurement object is calculated by applying a plurality of fringe images to the equation of the phase shift method, and the phase distribution is coded The difference between the unwrapping means for unwrapping based on the image, the unwrapped image that is the phase distribution after unwrapping, and the smoothed image of the unwrapped image is obtained, and the difference among the unwrapped images is obtained. Pixel but larger than the threshold value and a determination means and inappropriate pixels.

  Further, one aspect of the arithmetic program illustrating the present invention is to operate a computer as one aspect of the arithmetic apparatus illustrating the present invention.

  Further, one aspect of the surface shape measuring apparatus exemplifying the present invention is to image a measurement object on which a fringe pattern having a sinusoidal luminance distribution is projected a plurality of times while changing the phase of the fringe pattern. A code pattern for identifying individual stripes included in the stripe pattern is projected onto the measurement object instead of the stripe pattern, and the measurement object is imaged in that state to capture the code image. A measurement apparatus to be acquired, and an aspect of an arithmetic apparatus of the present invention that processes a fringe image and a code image acquired by the measurement apparatus.

  Further, in one aspect of the surface shape measuring method illustrating the present invention, a plurality of images of a measurement object on which a fringe pattern having a sinusoidal luminance distribution is projected are imaged a plurality of times while changing the phase of the fringe pattern. A code pattern for identifying individual stripes included in the stripe pattern is projected onto the measurement object instead of the stripe pattern, and the measurement object is imaged in that state to capture the code image. By applying the measurement procedure to be acquired and multiple fringe images to the phase shift method equation, the phase distribution of the fringe pattern deformed according to the surface shape of the measurement object is calculated, and the phase distribution is unwrapped based on the code image The difference between the unwrapping procedure to be performed, the unwrapped image that is the phase distribution after unwrapping, and the smoothed image of the unwrapped image is obtained. The off pixels and a determining procedure inappropriate pixel.

  According to the present invention, in the surface shape measurement by the phase shift method, it is possible to reliably detect an unwrapped failure location with respect to the phase image.

Configuration diagram of surface shape measuring device Operation flowchart of CPU 15 at the time of measurement Diagram explaining sine wave pattern Diagram explaining rectangular grid pattern Operation flowchart of CPU 15 at the time of analysis Schematic diagram of a map derived from the code image _I 5, _{I 6} and the code image _I 5, _{I 6,} Schematic diagram explaining the correction process

[First Embodiment]
Hereinafter, a first embodiment of the present invention will be described. This embodiment is an embodiment of a surface shape measuring apparatus.

  FIG. 1 is a configuration diagram of a surface shape measuring apparatus according to the present embodiment. As shown in FIG. 1, the surface shape measurement apparatus includes a stage 12 on which a measurement object 11 is placed, a projection unit 13, an imaging unit 14, a CPU 15, a storage unit 16, a monitor 17, and an input unit 18. have. Of these, the projection unit 13, the imaging unit 14, the storage unit 16, the monitor 17, and the input unit 18 are each connected to the CPU 15.

  The projection unit 13 is arranged in such a posture that the optical axis is inclined with respect to the reference surface (the surface on which the measurement object 11 is placed) of the stage 12, and faces the measurement object 11 of the stage 12. Irradiate pattern light. The projection unit 13 includes a light source 21, an illumination optical system 22, a pattern formation unit 23, and a projection optical system 24.

  The illumination optical system 22 converts the illumination light from the light source 21 into a parallel light beam, and illuminates the pattern forming unit 23 with the light beam. The pattern forming unit 23 is a panel (liquid crystal display element or the like) having a variable transmittance distribution, and gives an intensity distribution to the parallel light flux from the illumination optical system 22.

  The pattern forming unit 23 has a striped pattern (sinusoidal lattice pattern) having a sinusoidal transmittance change in one direction, a striped pattern (rectangular lattice pattern) having a rectangular wave shaped transmittance change in the same direction, and the like. Displayed as appropriate. Further, the phase of the sine lattice pattern displayed on the pattern forming unit 23 can be changed, and the pitch of the rectangular lattice pattern displayed on the pattern forming unit 23 can also be changed.

  The light beam that has passed through the pattern forming unit 23 enters the measurement object 11 on the stage 12 via the projection optical system 24, and thereby a pattern (sine lattice pattern or rectangular lattice pattern) is projected onto the measurement object 11. The Since the pattern projected on the measurement object 11 is deformed according to the surface shape of the measurement object 11, the surface shape of the measurement object 11 can be made known from the deformation amount.

  The imaging unit 14 is arranged so that the optical axis is perpendicular to the reference plane of the stage 12 and images the measurement object 11 on which the pattern is projected. The imaging unit 14 includes an imaging optical system 25 that forms an image of reflected light from the measurement object 11 and an imaging element 26 that captures an image of the measurement object 11 formed by the imaging optical system 25. Yes.

  The CPU 15 is a processor that performs measurement by comprehensively controlling the above-described units, and takes in and analyzes information acquired by the measurement.

  At the time of measurement, the CPU 15 operates the projection unit 13 and the imaging unit 14 to acquire an image of the measurement object 11 on which the pattern is projected. The CPU 15 also controls the pattern forming unit 23 to switch the pattern displayed on the pattern forming unit 3 between a sine lattice pattern and a rectangular lattice pattern, shift the phase of the sine lattice pattern, The image acquisition conditions can be variously changed by switching the pitch of the image.

  At the time of analysis, the CPU 15 obtains surface shape data of the measurement object 11 by calculating image data obtained by the imaging unit 14 under various conditions.

  The storage unit 16 is configured by a nonvolatile storage medium such as a flash memory. The storage unit 16 can store image data captured by the imaging unit 14 and surface shape data acquired by the CPU 15. The storage unit 16 stores an operation program for the CPU 15, and the CPU 15 operates according to the program.

  The monitor 17 displays an image captured by the image capturing unit 14 in accordance with an instruction from the CPU 15, a surface shape acquired by the CPU 15, and the like. The input unit 18 receives various inputs including a measurement start instruction from the user.

  FIG. 2 is an operation flowchart of the CPU 15 at the time of measurement. Hereinafter, each step will be described in order.

  Step S1: The CPU 15 sets a pattern to be displayed on the pattern forming unit 3 to a sine lattice pattern as shown in FIG. The luminance distribution in the lattice pitch direction of this sine lattice pattern is as shown in FIG. Hereinafter, it is assumed that the number of sine waves included in the sine lattice pattern is four.

Step S2: The CPU 15 repeats the imaging of the measuring object 11 four times while changing the phase of the sine lattice pattern displayed on the pattern forming unit 3 by π / 2 over one period, and the four striped images I ₁ , When I ₂ , I ₃ , and I ₄ are acquired, the fringe images I ₁ , I ₂ , I ₃ , and I ₄ are temporarily stored in the storage unit 16. These fringe images I ₁ , I ₂ , I ₃ , and I ₄ are used in phase calculation described later.

  Step S3: The CPU 15 sets a pattern to be displayed on the pattern forming unit 3 to a rectangular lattice pattern as shown in FIG.

Step S4: The CPU 15 switches the pattern displayed on the pattern forming unit 3 to a rectangular lattice pattern as shown in FIG. 4D, and images the measurement object 11 before and after the switching. The code images I ₅ and I ₆ are obtained. These code images I ₅ and I ₆ are temporarily stored in the storage unit 16.

  Here, the rectangular lattice patterns shown in FIGS. 4C and 4D are patterns for identifying individual stripes included in the sine lattice pattern shown in FIG. 3A, and are used for unwrapping. This pattern is necessary to make the distribution of the offset number n to be known on the image known. The rectangular lattice patterns shown in FIGS. 4C and 4D are obtained by the measurer by, for example, the following procedures (a) to (c).

  (A) Identification numbers N = 0, 1, 2, and 3 are assigned to individual stripes (individual sine waves) included in the sine lattice pattern shown in FIG. FIG. 4A shows a mapping of this identification number N.

  (B) Each identification number N in the map shown in FIG. 4 (A) is binary-coded as shown in FIG. 4 (B). As a result, the identification numbers N = 0, 1, 2, 3 are converted into the identification numbers N ′ = 00, 01, 10, 11 displayed in binary.

  (C) The bits of each identification number N ′ shown in FIG. 4B are allocated to different maps (here, two maps), and each bit is represented by a black or white pattern. Here, “1” is represented by a black pattern and “0” is represented by a white pattern. As a result, the two rectangular lattice patterns shown in FIGS. 4C and 4D are obtained.

  FIG. 5 is an operation flowchart of the CPU 15 at the time of analysis. Hereinafter, each step will be described in order.

Step S10: The CPU 15 reads out the stripe images I ₁ , I ₂ , I ₃ , and I ₄ from the storage unit 16 and performs phase calculation.

Specifically, CPU 15 may apply fringe images _{I 1, I 2, I 3} , the pixel value of the pixel number i included in the _{I 4 I 1i, I 2i,} I 3i, the expression of the following four bucket method _{I 4i} Thus, the initial phase φ _i at the pixel number _i is calculated.

Then, the CPU 15 performs the calculation of the initial phase φ _i for all the pixel numbers i, and the distribution of the initial phase φ _i obtained thereby is used as the phase image φ reflecting the surface shape of the measurement object 11. The data is temporarily stored in the storage unit 16.

Step S11: The CPU 15 reads the code images I ₅ and I ₆ from the storage unit 16. As shown in FIGS. 6A and 6B, the code images I ₅ and I ₆ show a state in which the rectangular lattice pattern shown in FIGS. 4C and 4D is deformed on the measurement object 11. . On these code images I ₅ and I ₆ , since the boundary part (edge) between the black pattern and the white pattern is dull, the SN ratio of the boundary part is lowered. The SN ratio is allowed to decrease, and the phase image φ is unwrapped by the following procedures (a) to (c).

(A) The CPU 15 refers to the pixel values I _5i and I _6i of the pixel number i included in the code images I ₅ and I ₆ , the color (white or black) indicated by the luminance value of the pixel value I _5i , and the pixel value From the combination with the color (white or black) indicated by the luminance value of I _6i , the optimum offset number n _i ′ (where n _i ′ is the number of offsets expressed in binary) is recognized. For example, when the combination is “black-black”, the offset number n _i ′ is “00”, and when the combination is “black-white”, the offset number n _i ′ is “01”. When the combination is “white-black”, the offset number n _i ′ is “10”, and when the combination is “white-white”, the offset number n _i ′ is “11”. It becomes. Then, the CPU 15 performs the calculation process of the offset number n _i ′ for all the pixel numbers i, and obtains a map of the offset number n _i ′ as shown in FIG.

(B) The CPU 15 converts each offset number n _i ′ shown in FIG. 6C into an offset number n _i displayed in decimal as shown in FIG. 6D.

(C) The CPU 15 applies the pixel value (phase φ _i ) of the pixel number i and the offset number n _i corresponding to the pixel number i on the map shown in FIG. 6D to the following equation (2). Thus, the pixel value after unwrapping (phase ψ _i ) is obtained. The CPU15 is this kind of unwrapping is performed for all the pixel number i, obtained phase [psi _i stores the distribution of the unwrapped image [psi as temporarily to the storage unit 16.

  Note that the unwrapped image ψ calculated in this step is caused by the decrease in the S / N ratio described above. For example, as shown in FIG. 7A, pixels a, b, c, d, e, f may be included.

  Step S12: When the CPU 15 obtains a smoothed image P by performing median filter processing on the unwrapped image ψ calculated in step S11, the CPU 15 temporarily stores the smoothed image P in the storage unit 16.

  Here, if the phase distribution of the unwrapped image ψ is as shown in FIG. 7A, the phase distribution of the smoothed image P is, for example, as shown in FIG. This smoothed image P is used as a reference when calculating an evaluation value.

  Step S13: The CPU 15 sets the pixel number i to be corrected to an initial value (1).

Step S14: The CPU 15 applies the pixel value (phase ψ _i ) of the pixel number i included in the unwrapped image ψ and the pixel value (phase P _i ) of the pixel number i included in the smoothed image P to the following expression. Thus, the evaluation value Δ _i of the pixel number _i is calculated.

Therefore, the relationship among the pixel value ψ _i , the pixel value P _i, and the evaluation value Δ _i is, for example, a relationship schematically shown in FIGS. 7A, 7B, and 7C.

Step S15: The CPU 15 compares the absolute value | Δ _i | of the evaluation value Δ _i calculated in step S14 with a threshold value T. The threshold T is set to a predetermined value between 0 and π as shown in FIG. 7C, and is, for example, π / 2. If the absolute value | Δ _i | is smaller, the CPU 15 regards the pixel with the pixel number i as an appropriate pixel and proceeds to step S21. If the absolute value | Δ _i | The process proceeds to step S16 considering that the pixel is appropriate.

  Therefore, if the pixel number i is any of the pixels a to f shown in FIG. 7C, the pixel number i is regarded as an inappropriate pixel in this step.

Step S16: CPU 15, the absolute value of the evaluation value delta _i | delta _i | is compared with [pi absolute value | delta _i | of it is regarded a pixel of the pixel number i and uncorrectable pixels smaller step S20 If the absolute value | Δ _i | is larger, the pixel with the pixel number i is regarded as a correctable pixel, and the process proceeds to step S17.

  Therefore, when the pixel number i is the pixel d shown in FIG. 7C, the pixel number i is regarded as an uncorrectable pixel in this step. On the other hand, when the pixel number i is one of the pixels a, b, c, e, and f shown in FIG. 7C, the pixel number i is a pixel that can be corrected in this step. It is regarded.

Step S17: CPU 15 refers to the sign of the evaluation value delta _i, the evaluation value delta _i is shifted to step S18 if it is positive, the evaluation value delta _i is shifted to step S19 if it is negative.

Step S18: The CPU 15 corrects the pixel value (phase ψ _i ) of the pixel number i by the following expression, and then returns to step S14 to determine whether or not the pixel is an appropriate pixel.

  Therefore, when the pixel with the pixel number i is one of the pixels a, c, and f shown in FIG. 7C, the pixel with the pixel number i is changed from FIG. 7C to (D) in this step. It is corrected as follows.

Step S19: The CPU 15 corrects the pixel value (phase ψ _i ) of the pixel number i by the following expression, and then returns to step S14 to determine whether or not the pixel is an appropriate pixel.

  Therefore, if the pixel number i is the pixels c and e shown in FIG. 7C, the pixel number i is corrected in this step as shown in FIGS. 7C to 7D. The

Step S20: The CPU 15 considers that the pixel value (phase ψ _i ) of the pixel number i is an error, and after replacing the phase ψ _i with a special value (for example, zero), proceeds to step S21.

  Therefore, when the pixel number i is the pixel d shown in FIG. 7C, the pixel number i is regarded as an error in this step and is not corrected at all.

  Step S21: The CPU 15 determines whether or not the above processing has been completed for all the pixel numbers i. If not completed, the process proceeds to step S23, and if completed, the process proceeds to step S22. To do.

  Step S22: The CPU 15 increments the pixel number i to be corrected by 1, and then returns to step S14.

Step S23: The CPU 15 visualizes the distribution of the pixel values (phase ψ _i ) of all the pixel numbers i on the monitor 17 as a corrected unwrapped image ψ. Thereby, the surface shape of the measuring object 11 becomes known. Further, the CPU 15 stores the corrected unwrapped image ψ in the storage unit 16 as necessary.

As described above, in steps S11 to S15 of the present embodiment, the individual pixel values (phase ψ _i ) of the unwrapped image ψ are compared with the corresponding pixel values (phase P _i ) of the smoothed image P, and evaluation showing the difference between the two values. If the value Δ _i is larger than the threshold value T, the pixel value (phase ψ _i ) is regarded as an inappropriate pixel.

Usually, the degree of error (unwrapping error) that occurs in the unwrapped image ψ due to unwrapping failure is sufficiently conspicuous compared to the degree of noise that occurs in the code images I ₅ and I ₆ . This is because the error generated in the unwrapped image ψ has a size corresponding to 2π (see pixels a, b, c, and e in FIG. 7A) or a size corresponding to an integer multiple of 2π. This is because there are many cases (see the pixel f in FIG. 7A).

  Therefore, in steps S11 to S15 of the present embodiment, inappropriate pixels (pixels a, b, c, d, e, and f) can be reliably found.

  In steps S18 and S19 of this embodiment, the pixel values (phases) of the inappropriate pixels (pixels a, b, c, e, and f) are corrected in units of 2π. ), The inappropriate pixels (pixels a, b, c, e) can be made appropriate pixels (pixels with sufficiently small evaluation value Δ).

In step S14~S19 of the present embodiment, since repeat the correction to evaluation value delta _i is less than the threshold T, greater unwrapped error as shown in FIG. 7 (C) → (D) → (E) An inappropriate pixel (pixel f) can also be made an appropriate pixel (a pixel having a sufficiently small evaluation value Δ).

  Further, in steps S16 and S20 of this embodiment, an inappropriate pixel (pixel d) having an evaluation value Δ between 0 and π is regarded as an error and correction is prohibited, so that the loop in steps S14 to S19 is unnecessary. (By the way, if the correction of such inappropriate pixels is not prohibited, the evaluation value Δ does not become small, and the loop in steps S14 to S19 is repeated infinitely. ).

[Supplement to the embodiment]
In the above-described embodiment, the processing target of steps S11 to S21 is the entire unwrapped image ψ, but the amount of calculation may be reduced by limiting to only a part of the unwrapped image ψ. This is because the probability of occurrence of an unwrapping error is high only in the vicinity of the boundary of the value n in the map shown in FIG.

  Therefore, in this case, the processing target of steps S11 to S21 may be limited to only the area A ′ (area corresponding to the vicinity of the map boundary) shown in FIG. 6D in the unwrapped image ψ. If the processing target is limited in this way, the amount of calculation can be efficiently reduced.

  In the above embodiment, the median filter process is performed on the unwrapped image ψ in acquiring the smoothed image P. However, another filter process (for example, a Gaussian filter process) having a smoothing function may be performed. However, it is considered that the median filter processing is more suitable than the Gaussian filter processing in order to suppress the influence of spike-like noise.

  In the above embodiment, the 4-bucket method in which the number of fringe images necessary for calculating the phase image φ is 4, but the 3-bucket method in which the number is 3 and the number is 7 is used. Other phase shift methods, such as some 7-bucket methods, may be applied.

  In the above embodiment, two patterns (binary code patterns) composed of two tones (white and black) are used as rectangular grid patterns for identifying individual stripes included in the sine grid pattern. One pattern (multi-color code pattern) consisting of two tones may be used.

  Moreover, in the said embodiment, although the number of the stripes contained in a sine lattice pattern was set to 4, it may increase more. However, in that case, since a larger number of identification numbers N is required, it is necessary to increase the number of binary code patterns (or the number of tones of a multicolor code pattern).

  In the above embodiment, the binary number display is used when encoding the identification number N. However, a gray code may be used.

  Further, the program stored in the storage unit 16 of the above embodiment may be a firmware program updated by version upgrade or the like. In other words, the measurement processing and analysis processing of this embodiment may be provided by updating the firmware program of the existing measurement processing and analysis processing.

  In the above embodiment, all of the measurement process shown in FIG. 2 and the analysis process shown in FIG. 5 are realized by software by the CPU 15, but all or part of the measurement process and analysis process is implemented by hardware by the ASIC. It may be realized as a wear.

DESCRIPTION OF SYMBOLS 11 ... Measuring object, 13 ... Projection part, 14 ... Imaging part, 15 ... CPU, 16 ... Memory | storage part, 17 ... Monitor, 18 ... Input part

US Pat. No. 6,075,605

Claims (10)

The measurement object on which the fringe pattern having a sinusoidal luminance distribution is projected is imaged a plurality of times while changing the phase of the fringe pattern to obtain a plurality of fringe images. The calculation device is applied to a measurement device that projects a code pattern for identifying stripes onto the measurement object instead of the stripe pattern and captures the measurement object in that state to obtain a code image. And
By applying the plurality of fringe images to the equation of the phase shift method, the phase distribution of the fringe pattern deformed according to the surface shape of the measurement object is calculated, and the unwrapping is performed to unwrap the phase distribution based on the code image. Means,
A determination unit that obtains a difference between the unwrapped image that is the phase distribution after unwrapping and a smoothed image of the unwrapped image, and determines a pixel in which the difference is greater than a threshold value as an inappropriate pixel,
An arithmetic device comprising: The arithmetic unit according to claim 1,
An arithmetic unit, further comprising a correction unit that corrects the value of the pixel determined by the determination unit as an inappropriate pixel in units of 2π. The arithmetic unit according to claim 2,
The correction means includes
The calculation device is characterized in that the correction is repeated until the pixel is not determined to be an inappropriate pixel. In the arithmetic unit according to claim 2 or 3,
The determination means includes
The threshold value is set between 0 and π. The arithmetic unit according to claim 4,
The correction means includes
Even if it is a pixel which the said determination means determined to be an improper pixel, the said correction | amendment is abbreviate | omitted to the pixel whose said difference was smaller than (pi). In the arithmetic unit according to any one of claims 1 to 5,
The determination means includes
An arithmetic device that uses, as the smoothed image, an image obtained by performing median filter processing on the unwrapped image before correction. In the arithmetic unit according to any one of claims 1 to 6,
The processing target of the determination means is
Of the unwrapped images, the calculation device is limited to only partial images corresponding to the boundary portions of the individual stripes.   An arithmetic program for causing a computer to operate as the arithmetic device according to claim 1. The measurement object on which the fringe pattern having a sinusoidal luminance distribution is projected is imaged a plurality of times while changing the phase of the fringe pattern to obtain a plurality of fringe images. A measurement device that projects a code pattern for identifying stripes onto the measurement object instead of the stripe pattern, captures the measurement object in that state, and acquires a code image;
The arithmetic device according to any one of claims 1 to 7, which processes a fringe image and a code image acquired by the measurement device;
A surface shape measuring device comprising: The measurement object on which the fringe pattern having a sinusoidal luminance distribution is projected is imaged a plurality of times while changing the phase of the fringe pattern to obtain a plurality of fringe images. A measurement procedure for projecting a code pattern for identifying stripes onto the measurement object instead of the stripe pattern and capturing the measurement object in that state to obtain a code image;
By applying the plurality of fringe images to the equation of the phase shift method, the phase distribution of the fringe pattern deformed according to the surface shape of the measurement object is calculated, and the unwrapping is performed to unwrap the phase distribution based on the code image. Procedure and
A determination procedure for determining a difference between the unwrapped image that is the phase distribution after unwrapping and a smoothed image of the unwrapped image, and determining a pixel in which the difference is greater than a threshold value as an inappropriate pixel,
A surface shape measuring method comprising:

Images