JP2010175554A - Device and method for measuring three-dimensional shape - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To rapidly and accurately measure the three-dimensional shape of an object. <P>SOLUTION: The three-dimensional shape measuring device includes: a projector 3 projecting each of a short-wavelength optical pattern 20 and long-wavelength optical pattern 30 onto the object 2; a camera 4 photographing the object 2 onto which the optical patterns 20 and 30 are projected; and a control part 5 which calculates relative phases of the optical patterns 20 and 30 based on an image photographed by the camera 4, determines an offset value between the relative phase and an absolute phase of the optical pattern 20 based on the calculated relative phase of the optical pattern 30, and determines the distance Z to the object 2 based on the offset value and the calculated relative phase of the optical pattern 20. In the device, the wavelength of the optical pattern 30 (long wavelength) is set at non-integer times the wavelength of the optical pattern 20 (short wavelength), and the correspondence relationship between the relative phase of the optical pattern 30 and the relative phase of the optical pattern 20 is made to be different for every period of the optical pattern 30. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a three-dimensional shape measurement apparatus and a three-dimensional shape measurement method, and more particularly to an apparatus and method that can measure a three-dimensional shape with high accuracy in a short time.

  Many methods for measuring the three-dimensional shape of an object in a non-contact manner have been proposed. As one of them, a sinusoidal grating phase shift method is known. This sine wave grating phase shift method is generally performed as follows.

As shown in FIG. 10, a light pattern having a sine wave luminance distribution is projected from a light source 72 to an object 73 through a grid 71 on which a light and shade pattern is printed in a sine wave pattern, and the stripe pattern on the object 73 is The image is taken by a camera 74 installed at a location different from the light source 72. An image is taken while shifting the grating 71 N times by 1 / N of the wavelength in the direction perpendicular to the stripes while the object 73 is stationary. For example, if the grating 71 is driven so that the phase of the fringe pattern is shifted by π / 2, and imaging is performed four times before the phase moves by one period, as shown in FIG. Four luminance values Ip ₍₀₎ , Ip ₍ π _{/ 2)} , Ip ₍ π ₎ , Ip ₍ 3π _{/ 2)} can be obtained for the measurement target point P inside. By substituting those luminance values into the following Equation 4, the phase φ of the light projected on the measurement target point P can be obtained. Here, (x, y) in Equation 4 is a coordinate value in the horizontal direction and the vertical direction of the measurement target point P.

  The obtained phase φ corresponds to the projection angle β of the light to the measurement target point P in order to indicate which part of the grating 71 the light reaching the measurement target point P passes as shown in FIG. Value. After the phase φ for the measurement target point P is obtained, the obtained phase value, the distance A between the camera 74 and the light source 72, and the photographing angle α when the object 73 is photographed are used according to the principle of triangulation. The distance Z to the measurement target point P is obtained. After that, the above processing is also performed for other measurement target points in the object 73 to obtain the distance Z, whereby the surface shape of the object 73 can be specified.

As described above, the phase φ of the light projected on the measurement target point P can be obtained using Equation 4, but since tan ⁻¹ is a periodic function having a period of 2π, the phase value derived by Equation 4 is used. φ can only take any value between 0 and 2π. For this reason, for example, even if the actual phase value φ is 7π / 2, the phase value φ derived from Equation 4 is 3π / 2, and the indefiniteness of an integral multiple of the period 2π remains.

Therefore, for example, in Patent Document 1, light (see FIG. 12) having a light and shade pattern obtained by adding a sine wave having a basic period and a sine wave having a period twice the basic period is projected onto an object. The phase value φ ₁ corresponding to the light of the light / dark pattern having the basic period and the phase value φ ₂ corresponding to the light of the light / dark pattern having the double period are obtained, and the phase value φ ₁ is referred to with reference to the phase value φ _2. There has been proposed a three-dimensional shape measuring apparatus that removes the indefiniteness of the phase value φ ₁ by connecting the phases.

Japanese Patent No. 3199041

  However, in the above three-dimensional shape measuring apparatus, the light to which the light and shade patterns of different periods are added is generated, and the light is projected onto the object to obtain the three-dimensional coordinates. Therefore, the distance to the measurement target point is obtained. However, there is a problem that an advanced mathematical method is required and the calculation processing time is prolonged.

  In the above three-dimensional shape measuring apparatus, when the distance to the measurement target point is obtained, the light projected on the object is converted into light corresponding to the fundamental period light and light having a period twice the fundamental period. Although it is necessary to divide into two corresponding parts, in order to correctly separate the two kinds of light, each light must have a correct sine wave characteristic. However, since the light projected from the light source has a non-linear effect due to lens distortion or the like, the waveform of the light of each period is distorted, and the distortion becomes an error in obtaining each light, and the measurement accuracy is deteriorated. There was a problem.

  Therefore, the present invention has been made in view of the problems in the conventional technology described above, and can reduce the calculation time and improve the measurement accuracy, and can accurately improve the three-dimensional shape in a short time. It is an object to provide a three-dimensional shape measuring apparatus and a three-dimensional shape measuring method capable of measuring the above.

  In order to achieve the above object, the present invention projects a sinusoidal light pattern on a measurement object, images the measurement object on which the light pattern is projected, and based on the captured image, A three-dimensional shape measurement apparatus for measuring a three-dimensional shape, wherein each of a first light pattern having a first wavelength and a second light pattern having a second wavelength longer than the first wavelength is used as the measurement object. Projecting means for projecting, photographing means for photographing a measurement object onto which the first light pattern and the second light pattern are projected, and the first light pattern and the first light pattern based on an image photographed by the photographing means. A phase calculating means for calculating the relative phase of the two light patterns, and an offset between the relative phase and the absolute phase of the first light pattern based on the relative phase of the second light pattern calculated by the phase calculating means. Phase connection means for obtaining a set value and obtaining a distance to the measurement object based on the offset value and the relative phase of the first light pattern calculated by the phase calculation means; and the second wavelength Is a length that is a non-integer multiple of the first wavelength, and the correspondence relationship between the relative phase of the second light pattern and the relative phase of the first light pattern is made different for each period of the second light pattern. To do.

  In the three-dimensional shape measuring apparatus, the first table in which the absolute phase of the first light pattern is registered in advance in association with the distance to the projectile of the first light pattern, and the relative phase of the second light pattern; A second table in which the offset value is registered in advance, and the phase connection unit collates the second table with the relative phase of the second light pattern calculated by the phase calculation unit, and An offset value is obtained, and an absolute phase of the first light pattern is obtained based on the offset value and a relative phase of the first light pattern calculated by the phase calculating means, and the absolute phase and the first table are obtained. And the distance to the measurement object can be obtained.

  Further, the present invention projects a sinusoidal light pattern onto the measurement object, images the measurement object on which the light pattern is projected, and measures the three-dimensional shape of the measurement object based on the captured image. A first step of projecting each of a first light pattern having a first wavelength and a second light pattern having a second wavelength longer than the first wavelength onto the measurement object. A second step of photographing the measurement object on which the first light pattern and the second light pattern are projected, and a relative phase between the first light pattern and the second light pattern based on the photographed image. Calculating an offset value between the relative phase and the absolute phase of the first light pattern based on the third step to be calculated and the relative phase of the second light pattern calculated in the third step; A fourth step of obtaining a distance to the measurement object based on the relative phase of the first light pattern calculated in the third step, and the second wavelength is a non-integer multiple of the first wavelength. The length of the second light pattern is different from the relative phase of the second light pattern and the relative phase of the first light pattern for each period of the second light pattern.

  As described above, according to the present invention, the calculation time can be shortened, the measurement accuracy can be improved, and the three-dimensional shape can be accurately measured in a short time.

1 is an overall configuration diagram showing an embodiment of a three-dimensional measurement apparatus according to the present invention. It is a block diagram which shows the control part of FIG. It is a schematic diagram which shows the offset table of FIG. It is a wave form diagram which shows the relative phase of a short wavelength, and the relative phase of a long wavelength. It is a flowchart explaining operation | movement of the three-dimensional measuring apparatus of FIG. It is a wave form diagram which shows the relative phase of a long wavelength. It is a wave form diagram which shows the relative phase of a short wavelength, and the relative phase of a long wavelength. It is a wave form diagram which shows the relative phase of a short wavelength. It is a wave form diagram which shows the relative phase of a short wavelength, and the relative phase of a long wavelength. It is a whole lineblock diagram showing an example of the conventional three-dimensional shape measuring device. It is a wave form diagram explaining operation | movement of the conventional three-dimensional shape measuring apparatus. It is a wave form diagram which shows the optical pattern used with the conventional three-dimensional shape measuring apparatus.

  Next, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is an overall configuration diagram showing an embodiment of a three-dimensional shape measuring apparatus 1 according to the present invention. The three-dimensional shape measuring apparatus 1 is a monocular three-dimensional shape measuring apparatus having one projection unit and one photographing unit, and transmits light having a predetermined density distribution (hereinafter referred to as “light pattern”) to the object 2. A projector 3 that projects the light, a camera 4 that captures the object 2 on which the light pattern is projected, and a control unit 5 that controls the projector 3 and the camera 4 and performs various arithmetic processes.

  The projector 3 includes a light source 3a that emits light having a predetermined intensity, and a sine wave grating 3b that displays a sine wave-like shading pattern and modulates the light from the light source 3a into a sine wave. The sine wave grating 3b is composed of, for example, a DMD (Digital Micro-mirror Device) that can control on / off of a reflection mirror in units of pixels, a liquid crystal panel that can control light transmittance in units of pixels, and the like, and a pattern of gray patterns It is possible to electronically change the interval, that is, the wavelength of the sine wave, or to electronically shift the phase of the shading pattern.

  In the present embodiment, when the light pattern is projected onto the object 2, the pattern interval of the shading pattern of the sine wave grating 3b is controlled, and a sine wave light pattern having a basic period (hereinafter referred to as “short wavelength light”). Two types of light patterns are projected: a “pattern” (20) and a sinusoidal light pattern (hereinafter referred to as a “long wavelength light pattern”) 30 having a period longer than the period of the light pattern 20. The long-wavelength light pattern 30 is set so that its period does not become an integral multiple of the period of the short-wavelength light pattern 20.

  Further, when projecting each light pattern of the short wavelength light pattern 20 and the long wavelength light pattern 30, the phase of the light and shade pattern displayed on the sine wave grating 3b is controlled so that the phase of each light pattern is π / Shifted by two.

  As shown in FIG. 2, the control unit 5 includes a camera control unit 10 that controls the operation of the camera 4, a projector control unit 11 that controls the operation of the projector 3, and an image capture that captures an image signal output from the camera 4. A storage unit 13 for storing an offset table, which will be described later, a phase calculation unit 14 for calculating a phase value, a phase connection unit 15 for obtaining a depth coordinate by connecting the calculated phase value, and a three-dimensional And a three-dimensional data generation unit 16 for generating data.

  The image capture unit 12 captures an image signal for one screen output from the camera 4 in synchronization with the phase shift operation of the projector 3 and stores the image signal in the storage unit 13. In the present embodiment, an image signal is captured at each timing when the phase shift amount of the light pattern becomes 0, π / 2, π, 3π / 2, 2π.

The phase calculation unit 14 calculates a phase value based on the image signal captured by the image capture unit 12. The phase value is calculated for each pixel when the short wavelength light pattern 20 is projected and when the long wavelength light pattern 30 is projected. The phase value φ ₁ when the short-wavelength light pattern 20 is projected is calculated by the following formula 1, and the phase value φ ₂ when the long-wavelength light pattern 30 is projected is calculated by the following formula 2. Is done.

Here, (x, y) are the horizontal and vertical coordinate values of the pixel, and I ₁ and I ₂ are the luminance values of the pixel. N ₁ and N ₂ are the number of phase shift shifts, and t ₁ and t ₂ are phase shift shift amounts.

Both of the phase values φ ₁ and φ ₂ calculated as described above have indefiniteness that is an integral multiple of the period 2π. In the following description, the phase value having indefiniteness is referred to as a “relative phase value”. The phase value from which the ambiguity is removed is referred to as “absolute phase value”. Also, the phase shifted short wavelength light pattern 20 is projected, the relative phase value obtained from the captured image is referred to as “short wavelength relative phase value”, and the phase shifted long wavelength light pattern 30 is projected. The relative phase value obtained from the captured image is referred to as a “long wavelength relative phase value”.

  The phase connection unit 15 obtains the offset value 2nπ and the absolute phase value ψ based on the relative phase value φ calculated by the phase calculation unit 14 and the offset table stored in the storage unit 13. The relationship between the offset value, the relative phase value φ, and the absolute phase value ψ is as shown in Equation 3 below. As is apparent from Equation 3, the offset value corresponds to the indefiniteness between the absolute phase and the relative phase. To do.

FIG. 3 is a schematic diagram illustrating an example of the offset table. As shown in FIG. 3, the offset table includes a first table 40 for a short wavelength light pattern and a second table 50 for a long wavelength light pattern. In the first table 40, as shown in FIG. 3A, the relative phase value φ ₁ of the short wavelength and the absolute value are associated with the distance Z ₁ from the reference point J (see FIG. 1) to the measurement target point P. The phase value ψ ₁ is registered for each pixel. On the other hand, in the second table 50, as shown in FIG. 3B, the relative phase value φ ₂ of the long wavelength and the short wavelength of the short wavelength are associated with the distance Z ₂ from the reference point J to the measurement target point P. The n value of the light pattern 20 is registered for each pixel.

Hereinafter, in order to distinguish the relative phase values φ ₁ , φ _2, etc. registered in the first table 40 and the second table 50 from the relative phase values φ ₁ , φ _2, etc. calculated at the time of measuring the three-dimensional shape, These are referred to as phase table values Tφ ₁ , Tφ _2, etc.

Each of the first table 40 and the second table 50 is created by sampling the relative phase value φ ₁ of the short wavelength and the relative phase value φ ₂ of the long wavelength before measuring the three-dimensional shape. It is. In creating the first table 40 and the second table 50, first, a short-wavelength optical pattern and a long wavelength are shifted while shifting the phase by π / 2 with a flat plate having a flat surface fixed at a predetermined reference position. Are projected onto a flat plate to obtain a relative phase value φ ₁ of a short wavelength and a relative phase value φ ₂ of a long wavelength, and are registered as relative phase table values Tφ ₁ and Tφ ₂ at a reference position.

Here, since the long wavelength light pattern has a longer period than the short wavelength light pattern, if the period of the short wavelength light pattern and the period of the long wavelength light pattern are known in advance, FIG. As shown, the offset value 2nπ of the short wavelength light pattern, that is, the n value can be derived from the long wavelength relative phase value φ ₂ . For example, when the long wavelength light pattern has a period of about 4.5 times that of the short wavelength light pattern, and the relative phase value φ ₂ of the long wavelength is π, the n value of the short wavelength light pattern is 2 can be derived. If the n value is known, the short wavelength absolute phase value ψ ₁ can also be obtained by adding the offset value 2nπ to the short wavelength relative phase value φ ₁ .

Therefore, in the present embodiment, after registering the relative phase table values Tφ ₁ and Tφ ₂ , the n value of the short-wavelength light pattern and the absolute phase value ψ ₁ of the short-wavelength light pattern are obtained using the above method. In addition, the obtained n value is registered in the second table 50 and the absolute phase value ψ ₁ is registered in the first table 40 as the absolute phase table value Tψ ₁ .

Thereafter, the flat plate is moved by a predetermined amount in the depth direction, and in the same manner as described above, the relative phase values φ ₁ , φ _2, etc. are obtained, and the obtained relative phase table is associated with the distances Z ₁ , Z ₂ The values Tφ ₁ , Tφ _2, etc. are registered in the first table 40 and the second table 50. Thereafter, the relative phase table values Tφ ₁ , Tφ _2, etc. at each position are sequentially registered while moving the flat plate in the depth direction by a predetermined movement amount, whereby the first table 40 and the second table 50 are stored. create. In the first table 40 and the second table 50, the distance to which various data such as the relative phase table values Tφ ₁ and Tφ ₂ are registered is set to the size of the measurement target range. Or, it can be appropriately determined according to the shape and size of the measurement object.

  Next, a method for measuring a three-dimensional shape will be described with reference to FIG.

  In measuring the three-dimensional shape of the object 2, as shown in FIG. 5, first, the light pattern 20 with a short wavelength is projected onto the object 2 with the phase shift amount of the light pattern being zero, and the camera 4 The output image signal is captured (step S1). Next, in a state where the phase of the light pattern is sequentially shifted by π / 2, π, 3π / 2, and 2π, the short-wavelength light pattern 20 is projected onto the object 2, and an image signal with each phase shift amount is captured. (Step S2).

Next, from the image signal captured when the phase shift amount of the light pattern is 0 and the image signal captured when the phase of the light pattern is shifted by π / 2, π, 3π / 2, 2π, the short wavelength The relative phase value φ ₁ when the light pattern 20 is projected is calculated (step S3). At this time, the relative phase value φ ₁ is calculated for each pixel.

Next, a light pattern 30 having a long wavelength is projected onto the object 2, and the phase of the image signal and the light pattern when the phase shift amount of the light pattern is set to 0 is shifted by π / 2, π, 3π / 2, and 2π. Each image signal is captured (steps S4 and S5). Then, based on each captured image signal, the relative phase value φ ₂ when the long-wavelength light pattern 30 is projected is calculated in units of pixels (step S6).

Next, one pixel is selected as a target pixel (measurement target point) from a plurality of pixels constituting one screen (step S7). Next, it is determined whether or not various data such as the relative phase table value Tφ ₂ are registered in the second table 50 created in advance more than one cycle of the long wavelength light pattern 30 (step S8). ).

If the data of more than one cycle is registered (step S8: Yes), the second table 50 is referred to, from among the distance Z ₂ registered in the second table 50, calculated in the step S6 The distance Z ₂ corresponding to the relative phase value φ ₂ is acquired (step S9). However, when data exceeding one period is registered, there may be two or more distances Z ₂ corresponding to the relative phase value φ ₂ as shown in FIG. A distance Z _2a closer to the reference point J is acquired as the distance Z ₂ .

Next, the first table 40 for the short wavelength light pattern is referred to, and the relative phase table value Tφ corresponding to the distance Z ₂ is selected from the short wavelength relative phase table values Tφ ₁ registered in the first table 40. ₁ is read (step S10). Then, the read relative phase table value Tφ ₁ is compared with the relative phase value φ ₁ calculated in the previous step S3 (step S11).

If there are two or more distances Z ₂ corresponding to the relative phase value φ ₂ , any one of them is a true distance, but if the distance Z ₂ acquired in step S9 is a true distance, FIG. As shown in (a), the relative phase table value Tφ ₁ and the calculated relative phase value φ ₁ should substantially match. Conversely, if the distance Z ₂ acquired in step S9 is not a true distance, the relative phase table value Tφ ₁ and the calculated relative phase value φ ₁ should not match as shown in FIG. 7B. Therefore, by comparing the read relative phase table value Tφ ₁ with the relative phase value φ ₁ calculated in the previous step S3, the distance Z ₂ acquired in step S9 is the true distance. It can be determined whether or not.

Result of the comparison, the relative phase table value Tifai _1, if the calculated relative phase values phi ₁ do not match (step S11: No), the distance Z ₂ obtained in step S9 is not true of the distance, A distance Z _2b far from the reference point J (see FIG. 6) is acquired as the distance Z ₂ (step S12). Thereafter, the process proceeds to step S10, and the comparison process between the relative phase table value Tφ ₁ and the relative phase value φ ₁ is executed in the same manner as described above.

On the other hand, when the relative phase table value Tφ ₁ and the calculated relative phase value φ ₁ substantially match (step S11: Yes), whether or not there are two or more distances Z ₂ that satisfy the condition. Is determined (step S13). When there are two or more distances Z ₂ that satisfy the condition, it is not possible to specify which one is correct. Therefore, the measurement of the three-dimensional shape for the pixel ends, and another pixel is selected ( Step S14).

On the contrary, if the condition is satisfied distance Z ₂ is one, the distance Z ₂ at that time because it is true distance, that distance Z ₂ is recognized as the final distance Z _2. Next, the second table 50 is referred to, and the n value corresponding to the distance Z ₂ is obtained from the n values of the short wavelength light patterns registered in the second table 50, and the offset value 2nπ is obtained. (Step S15).

On the other hand, the second table 50, in the case where only one period following the data of the light pattern 30 having a long wavelength is not registered (step S8: No), the distance Z ₂ corresponding to the relative phase values phi ₂ calculated because but there is only one, the distance Z ₂ corresponding to the relative phase values phi ₂ based on the second table 50 is acquired (step S16), and then, in the same manner as described above, the offset value 2nπ is obtained (step S15).

When the offset value 2nπ is obtained, the obtained offset value 2nπ is added to the short wavelength relative phase value φ ₁ calculated in step S3 to obtain the absolute phase value ψ ₁ of the short wavelength light pattern (step _1). S17). Next, it is determined whether or not the short wavelength relative phase value φ ₁ is 0, 2π, or a value in the vicinity thereof (step S18). Here, the value in the vicinity of 0, 2π means a value where the difference from 0 or 2π is about 0.2π to 0.3π. Specifically, the relative phase value φ ₁ = 0 to 0.3π. 1.7π-2π.

Here, in step S18, to carry out the above determination process, when the relative phase values phi ₁ short wavelength is a value 0,2π or near thereof, the error is acquired offset value 2nπ in step S15 Because there is a possibility. That is, when the phase is plotted on the vertical axis and the distance is plotted on the horizontal axis and the characteristics of the relative phase are graphed, theoretically, as shown in FIG. 8A, the length L ₁ in the horizontal axis direction, A waveform in which right triangles having the same L ₂ and L ₃ are continuous should be obtained. However, in practice, the waveform as shown in FIG. 8A is not always obtained, and is affected by fluctuations in light intensity, projection angle, etc., as shown in FIG. There is a case where a waveform in which right-angled triangles having irregular lengths L ₁ ′, L ₂ ′, and L ₃ ′ in the horizontal axis direction are continuous is obtained. On the other hand, the offset value (n value) registered in the second table 50 is not an actual measurement value, but is based on the premise that the relative phase of the short-wavelength light pattern has theoretical characteristics (FIG. 8A). This is the theoretical value derived by. Therefore, for example, as shown in FIG. 9, if the relative phase value φ ₁ of the short wavelength is a value in the vicinity of 2π in a state where the actual L ₁ ′ is longer than the theoretical L ₁ , Actually, although the n value = 0, the offset value 2nπ is obtained with the n value = 1. As a result, as the distance to the surface of the object 2 (distance Z ₁ ), the distance of the H point whose distance is increased by 2π although the distance of the G point in FIG. 9 should be obtained. It will be calculated as the distance.

Step S18 is a process for preventing such an error, and when the relative phase value φ ₁ is 0, 2π or a value in the vicinity thereof (step S18: Yes), it is registered in the first table 40. The distance Z ₁ corresponding to the absolute phase value ψ ₁ is obtained from the distance Z ₁ being obtained (step S19), the obtained distance Z ₁ and the distance Z ₂ obtained in steps S8 to S16 Are compared, and it is determined whether or not both have substantially the same value (step S20).

Upon comparing the distance Z ₁ and the distance Z _2, rather than whether they match, to determine whether the approximately same value, the distance Z ₁ is based on the short-wavelength light pattern The distance Z ₂ is a value in the second table 50 created based on the long wavelength light pattern, and the short wavelength light pattern and the long wavelength. This is because there is a possibility that a difference of about several millimeters may occur between the two even if the offset value 2nπ is a correct value.

If the distance Z ₁ and the distance Z ₂ have substantially the same value (step S20: Yes), the offset value 2nπ acquired in step S15 can be estimated to be a correct value, and therefore the distance Z _{1 at} that time Is recognized as the depth coordinate of the measurement target point.

On the other hand, when the distance Z ₁ and the distance Z ₂ have completely different values (step S20: No), the n value of the offset value 2nπ acquired in step S15 indicates that the n value is incorrect. ± 1 and the offset value 2nπ is changed (step S21). Then, the process of step S17~ step S20 is executed, through the determination process of step S18 and step S20, the correct offset values 2nπ and distance Z ₁ is derived, the depth coordinates of the measurement target point is obtained.

On the other hand, if the relative phase value φ ₁ is not 0, 2π, or a value in the vicinity thereof (step S18: No), it can be inferred that the offset value 2nπ acquired in step S15 is a correct value. Similarly, the distance Z ₁ corresponding to the absolute phase value ψ ₁ is acquired, and the distance Z ₁ is recognized as the depth coordinate of the measurement target point (step S22).

  When the depth coordinates of the measurement target point are acquired, it is determined whether or not pixels (measurement target points) for which the depth coordinates have not been acquired remain (step S23). If there is a pixel for which the depth coordinate has not been acquired, that pixel is selected (step S24), and the depth coordinate is acquired in the same manner as described above (steps S8 to S22). Thereafter, the processing in steps S8 to S22 is repeatedly executed until there are no pixels for which the depth coordinates have not been acquired, and the depth coordinates are acquired for all measurement target points on the surface of the object 2. Finally, the acquired depth coordinates and the vertical and horizontal coordinates of each measurement target point are integrated, and three-dimensional data indicating the three-dimensional shape of the object 2 is generated (step S25).

As described above, according to the present embodiment, the offset value 2nπ is obtained from the relative phase value φ ₂ obtained by projecting the long wavelength light pattern 30, and the offset value 2nπ and the short wavelength light pattern are obtained. In order to obtain the depth coordinate of the measurement target point from the relative phase value φ ₁ obtained by projecting 20, an arithmetic process is performed when obtaining the short wavelength relative phase value φ ₁ and the long wavelength relative phase value φ _2. Although it is necessary, if only the relative phase values φ ₁ and φ _{2 are} calculated, then the depth coordinates of the measurement target point can be obtained by a simple process. Therefore, the calculation time can be shortened, and the three-dimensional shape of the object 2 can be measured quickly.

  In addition, according to the present embodiment, the shape of the object 2 is mainly high-resolution short-wavelength light while separately using two types of light patterns, the short-wavelength light pattern 20 and the long-wavelength light pattern 30. Since the measurement is performed using the pattern 20 and the offset value 2nπ at that time is measured using the long wavelength light pattern 30 having a large dynamic range, the measurement accuracy can be improved and the three-dimensional shape can be measured with high accuracy. Is possible.

Furthermore, according to the present embodiment, the relative relationship between the first table 40 in which the absolute phase value ψ ₁ of the short wavelength light pattern 20 is registered in association with the distance Z ₁ to the object 2 and the long wavelength light pattern 30. The second table 50 in which the offset value 2nπ (n value) is registered in association with the phase value φ ₂ is created, and using them, the depth coordinates of the measurement target point are obtained, so that the offset value 2nπ and The depth coordinates of the measurement target point can be obtained.

  The embodiment of the present invention has been described above, but the present invention is not limited to the above-described configuration, and various modifications can be made within the scope of the invention described in the claims.

For example, in the above embodiment, when the offset value 2nπ is acquired, the n value is obtained from the distance Z ₂ with reference to the second table 50 (FIG. 5: step S15), but the long value instead of the distance Z ₂ is obtained. The wavelength relative phase value φ ₂ may be used to obtain the n value from the long wavelength relative phase value φ ₂ .

Further, in the above embodiment, the true / false of the offset value 2nπ is determined by determining whether or not the relative phase value φ ₁ is 0, 2π or a value in the vicinity thereof (FIG. 6: step S18). However, the absolute phase value ψ ₁ is used in place of the relative phase value φ _1, and it is determined whether or not the short wavelength absolute phase value ψ ₁ is an integer multiple of 0, 2π or a value in the vicinity thereof. Thus, the authenticity of the offset value 2nπ may be determined.

DESCRIPTION OF SYMBOLS 1 3D shape measuring device 2 Object 3 Projector 3a Light source 3b Sine wave grating 4 Camera 5 Control part 10 Camera control part 11 Projector control part 12 Image capture part 13 Storage part 14 Phase calculation part 15 Phase connection part 16 Three-dimensional data generation Section 20 Short Wave Pattern 30 Long Wave Pattern 40 First Table 50 Second Table

Claims (3)

A three-dimensional shape measuring apparatus that projects a sinusoidal light pattern onto a measurement object, images the measurement object on which the light pattern is projected, and measures the three-dimensional shape of the measurement object based on the captured image Because
Projection means for projecting each of a first light pattern having a first wavelength and a second light pattern having a second wavelength longer than the first wavelength onto the measurement object;
Imaging means for imaging the measurement object on which the first light pattern and the second light pattern are projected,
Phase calculating means for calculating a relative phase between the first light pattern and the second light pattern based on an image photographed by the photographing means;
Based on the relative phase of the second light pattern calculated by the phase calculation means, an offset value between the relative phase and the absolute phase of the first light pattern is obtained, and the offset value and the phase calculation means are calculated. Phase connection means for obtaining a distance to the measurement object based on the relative phase of the first light pattern,
The length of the second wavelength is a non-integer multiple of the first wavelength, and the correspondence relationship between the relative phase of the second light pattern and the relative phase of the first light pattern is varied for each period of the second light pattern. A three-dimensional shape measuring apparatus characterized by that. A first table in which the absolute phase of the first light pattern is registered in advance in association with the distance to the projectile of the first light pattern;
A second table in which the offset value is registered in advance in association with the relative phase of the second light pattern;
The phase connecting means obtains the offset value by collating the relative phase of the second light pattern calculated by the phase calculating means with the second table, and is calculated by the offset value and the phase calculating means. The absolute phase of the first light pattern is obtained based on the relative phase of the first light pattern, and the distance to the measurement object is obtained by comparing the absolute phase with the first table. The three-dimensional shape measuring apparatus according to claim 1. A three-dimensional shape measurement method for projecting a sinusoidal light pattern onto a measurement object, photographing the measurement object on which the light pattern is projected, and measuring the three-dimensional shape of the measurement object based on the photographed image Because
A first step of projecting each of a first light pattern having a first wavelength and a second light pattern having a second wavelength longer than the first wavelength onto the measurement object;
A second step of photographing the measurement object on which the first light pattern and the second light pattern are projected;
A third step of calculating a relative phase of the first light pattern and the second light pattern based on a photographed image;
An offset value between the relative phase and the absolute phase of the first light pattern is obtained based on the relative phase of the second light pattern calculated in the third step, and the offset value is calculated in the third step. A fourth step of determining a distance to the measurement object based on the relative phase of the first light pattern;
The length of the second wavelength is a non-integer multiple of the first wavelength, and the correspondence relationship between the relative phase of the second light pattern and the relative phase of the first light pattern is varied for each period of the second light pattern. A three-dimensional shape measuring method characterized by that.

Images