JP3178436B2 - Phase connection method and apparatus, and recording medium - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a phase connecting method and apparatus, and more particularly, to a phase connecting method and apparatus for determining a connection relation of phases calculated by a phase shift method in three-dimensional shape measurement, and a recording medium.

[0002]

2. Description of the Related Art As one of various methods for measuring a three-dimensional shape from an image, a grid pattern projection method has been conventionally known. For example, reference is made to the description of this grid pattern projection method in the literature (“Lattice pattern projection method introducing fringe scanning”, Journal of the Precision Society, pp. 1817-1822, 1989).

FIG. 2 is a diagram showing a schematic configuration of a three-dimensional shape measuring apparatus described in the above-mentioned document. Referring to FIG. 2, a sine-wave grating pattern is projected from a light source 101 through a grating 102 onto an object 100. It is assumed that the grid 102 has gray-scale values printed in a sine wave shape. Point P10 of object 100
Assuming that the coordinate value on the image of the point P104 is x and the luminance value at x is I (x) when 4 is photographed by the camera 103, the luminance value can be expressed by the following equation (1).

I (x) = a ₀ (x) + A (x) cos (φ + α) (1)

In equation (1), a ₀ (x) is a bias component, and φ and α are phases.

Here, φ is set while the object 100 is stationary.
2 (N-1) π in increments of 2π / N radians from 0 radians
/ N radians (where N is a positive integer), and an image is taken by the camera 103 each time it is shifted. This operation can be realized by taking an image while shifting the grating 102 N times in the direction of the axis u by 1 / N pitch of the wavelength of the printed sine wave pattern.

The luminance value I _k (x) at the point x obtained at the k-th time (0 ≦ k ≦ N) can be expressed by the following equation (2).

I _k (x) = a ₀ (x) + A (x) cos (2πk / N + α) (2)

After all, the phase α is a phase value at the point x in the image captured at the k = 0th position. When viewing the point P104 on the object 100 from the camera 103, it can be seen that there is a point P104 in the direction of a straight line passing through the lens center from the coordinates x on the camera screen.

When viewed from the light source 101, it can be seen that a point P104 is on a plane defined by a straight line on the sine wave grating 102 having a phase α at k = 0 and the center of the light source 101. Therefore, it is possible to calculate the intersection of the straight line and the plane as the position of the point P104 in the three-dimensional space.

The above equation ( 2 ) is obtained by introducing two new coefficients a ₁ (x) and b ₁ (x) as shown in the following equation (3). Can be rewritten as

A ₁ (x) = A (x) cos (α) b ₁ (x) = − A (x) sin (α) (3)

I _k (x) = a ₀ (x) + a ₁ (x) cos (2πk / N) + b ₁ (x) sin (2πk / N) (4)

Here, a ₁ (x) and b ₁ (x) are luminance values I ₀ (x),.
_{From N-1} (x), it can be obtained by the following equation (5). Therefore, the phase α can be obtained by the following equation (6).

[0015]

By performing the above calculation on the image data captured by the camera 103, the phase α can be obtained on the object 100 on the image. Further, based on the mutual positional relationship between the light source 101, the grid 102, and the camera 103, the object 1
00 three-dimensional shapes can be obtained.

By the way, as for α obtained from the above equation (6), the function tan ⁻¹ is simply applied, so that the obtained phase α has an uncertainty of an integral multiple of 2π.

This problem can be avoided by projecting a sine wave pattern consisting of only one period onto the object. However, in order to assign an initial phase having a narrow range of -π to π to the entire object, However, there is a problem that the measurement accuracy cannot be increased.

For this reason, as described in the above-mentioned document, a method of projecting a sine wave pattern having a plurality of periods to expand the value range of the initial phase and improve the accuracy is usually adopted. FIG. 3 shows an example of the phase image 105 obtained by calculating the phase of each pixel on the image captured by the camera 103 in this case. In FIG. 3, the range from black to white is -π to π radian. As is clear from FIG. 3, α obtained by the above equation (6) can be obtained only with a value from −π to π, and the initial phase becomes indefinite by an integral multiple of 2π. For this reason, even if a plane is measured, for example, a discontinuous phase is obtained, and it is necessary to reconstruct a plane phase by some method.

In the above-mentioned document, the problem of phase uncertainty is eliminated by a simple phase connection method. That is, when the phase between adjacent pixels changes by about 80% of 2π, 2π is added or subtracted so that the difference is minimized, thereby converting the phase into a continuous phase.

[0021]

As described above, in the conventional phase connection method, smoothing is performed between adjacent pixels, but in many cases, the movement does not actually work well. It suffices if there are only a pair of adjacent pixels, but since the image has a two-dimensional spread, depending on the relationship between the upper, lower, left and right pixels, noise, the complicated shape of the target object, the boundary between the background and the object For example, a situation occurs in which contradictions occur and connection cannot be made smoothly.

In this case, the influence of noise can be reduced by increasing the number of times of image capturing. However, additional processing time is required to reduce noise due to shooting time, image storage area, averaging operation, and the like.

The influence of the complicated shape of the target object can be reduced by projecting a sine wave pattern having a low spatial frequency, but there is a problem that the measurement accuracy is reduced. In this case, it is possible to simultaneously project a high-frequency sine wave pattern and smoothly connect the phases while referring to the continuous phase value obtained from the low-frequency sine wave pattern. However, there is a problem that the area for saving the data and the processing time become longer.

Accordingly, the present invention has been made in view of the above problems, and has as its object to provide a phase connection device and a method for accurately and quickly determining a connection relationship of phases calculated by a phase shift method. To provide.

[0025]

To achieve the above object, the first invention of the present application is as follows. (A) A phase image obtained by a phase shift method is inputted, and a predetermined size including a pixel of interest in the phase image is input. Setting a test area, and determining that the phase of the target pixel can be connected when all the differences between the phase of the target pixel and the phases of the pixels in the test area are smaller than a predetermined threshold value; b) for all pixels of the phase image, (a)
(C) selecting one pixel from the group of phase-connectable pixels obtained in the steps (b) and (b), and setting the phase value of the pixel to a definite phase value. And (d) selecting a pixel whose phase is not determined for a group of pixels other than the one pixel that can be connected in phase and whose adjacent pixels have a determined phase value, a step difference of the possible pixel and the definite phase value is a definite phase value by adjusting the integral multiple of 2π radians to be less than 2π radians, the process steps of (e) above (d), the phase connection Repeating until all possible pixel groups have a deterministic phase value.

According to a second aspect of the present invention, (a) a phase image obtained by a phase shift method is input, a test area of a predetermined size including a pixel of interest is set in the phase image, and the phase of the pixel of interest and the A step of determining that the target pixel can be connected in phase when all of the phase differences of the pixels in the test area are smaller than a predetermined threshold value; and a) performing the phase connection feasibility determination process; and (c) when the group of phase connectable pixels obtained in the steps up to (b) constitutes a plurality of continuous regions, Selecting a continuous pixel group;
(D) selecting one pixel from the selected group of phase-connectable pixels and setting the phase value of the pixel to a definite phase value; and (e) selecting the other selected pixel other than the one pixel. The phase is not determined for the group of pixels that can be connected in phase, and an adjacent pixel selects a pixel having a determined phase value, and the difference between the other phase-connectable pixels and the determined phase value is less than 2π radians. Adding and subtracting an integral multiple of 2π radians to obtain a definite phase value so as to reduce the value; and (f) repeating the processing of the step (e) until all the selected continuous pixel groups have a definite phase value. And (g) the above (c) to (f)
Is repeated until all the phase-connectable pixel groups have the determined phase value.

According to the third invention of the present application, (a) a phase image obtained by a phase shift method is input, a test area of a predetermined size including a pixel of interest is set in the phase image, and the phase of the pixel of interest and When all of the phase differences of the pixels in the test area are smaller than a predetermined threshold value, a step of determining that the pixel of interest can be connected in phase; Performing a phase connection possibility determination process;
(C) a step of selecting one continuous pixel group from one or a plurality of continuous pixel groups designated when the plurality of phase-connectable pixel groups constitute a plurality of continuous regions; and (d) selecting the selected continuous pixel group. Selecting one pixel from the group of phase-connectable pixels and setting the phase value of the pixel to a definite phase value; and (e) selecting a pixel group other than the one pixel that can be connected to the selected phase. On the other hand, an integer of 2π radian is selected such that a pixel whose adjacent phase does not have a fixed phase value and whose adjacent phase has a fixed phase value is smaller than 2π radian. (F) repeating the processing of the above steps (d) to (e) until all the selected continuous pixel groups have a confirmed phase value, g) One or more designated steps (c) to (f) Repeating until a plurality of phase-connectable pixel groups have a definite phase value.

The fourth invention of the present application is the invention according to the second invention or the third invention, wherein (h) the deterministic phase value is input, and one continuous region is connected to a plurality of phase-connectable continuous regions having the deterministic phase value. And selecting a phase value group of the selected continuous region as a globally defined phase value group; and (i) a continuous region having a globally defined phase value group among continuous regions having no globally defined phase value group. Selecting one continuous region closest to the region; and (j) applying 2π to the phase value group of the continuous region that does not have the wide-area fixed phase value group so that the phase value group of both continuous regions is connected most smoothly. And (k) repeating the above steps (h) to (j) until all continuous regions take a group of globally defined phase values.

In a fifth aspect of the present invention, in any one of the first to third aspects of the present invention, (l) inputting the deterministic phase value, and determining the deterministic phase value in a pixel group determined to be difficult to phase-connect with the deterministic phase value And (m) adding or subtracting an integral multiple of 2π radians such that the phase difference between the estimated phase value and the pixel determined to be difficult to phase-connect is minimized.
(N) repeating the above steps (l) and (m) until all the pixel groups determined to be difficult to phase-connect have a definite phase value.

The sixth invention of the present application is the fourth invention, wherein
(O) inputting the globally determined phase value and estimating a globally determined phase value from the globally determined phase value in a pixel group determined to be difficult to phase-connect; Adding or subtracting an integral multiple of 2π radians so that the phase difference of the determined pixel is minimized;
(Q) a step of repeating the above steps (o) and (p) until all the pixel groups determined to be difficult to phase-connect have a globally defined phase value.

According to a seventh aspect of the present invention, a phase image calculated by a phase shift method, a test area size signal, and a threshold signal are inputted, and a test area of a predetermined size including a pixel of interest in the phase image is input. When the difference between the phase of the pixel of interest and the phase of the pixel in the test area is all smaller than the predetermined threshold value, a phase connection possibility determination process for determining that the pixel of interest is phase connectable is performed. A phase connection test unit that executes the phase connection possibility determination process for all the pixels in the phase image and outputs a mask image composed of the phase connectable pixels; and the phase image and the mask. An image is input, one pixel is selected from a group of pixels that can be connected in phase, and the phase value of the pixel is set as a definite phase value. Have value 2 so that the difference between the yet-definite phase value as determined phase value by selecting the pixel is less than 2π radians
phase connection means for acquiring and outputting a connected phase image by adding and subtracting an integral multiple of π radians to obtain a definite phase value, and repeating the above processing until all of the pixels capable of phase connection have a definite phase value. .

According to an eighth aspect of the present invention, a phase image calculated by the phase shift method, a test area size signal, and a threshold signal are input, and a test area of a predetermined size including a pixel of interest in the phase image. When the difference between the phase of the pixel of interest and the phase of the pixel in the test area are all smaller than a predetermined threshold, a phase connection determination process is performed to determine that the pixel of interest can be connected to a phase. Phase connection test means for performing the phase connection possibility determination process on all the pixels of the phase image and outputting a mask image composed of the phase connectable pixels; Is configured, the phase image and the mask image are input, one continuous pixel group is selected, one pixel in the continuous area is selected, and the phase value of the pixel is set as a definite phase value, Rank from group Is not determined, and the adjacent pixel selects a pixel having a determined phase value, and adds and subtracts an integral multiple of 2π radians so that the difference between the undetermined phase value and the determined phase value is smaller than 2π radians. Phase connection means for acquiring and outputting a connected phase image by repeating the above processing until all of the continuous pixel groups have a definite phase value, and repeating the above processing until all of the continuous regions have a definite phase value And

According to a ninth aspect of the present invention, in the above-mentioned eighth aspect, when a group of pixels that can be connected in phase forms a plurality of continuous areas, a mask image is input and one of the continuous areas is selected to select a mask. Continuous area selecting means for outputting an image,
Phase connection means for obtaining a fixed value of the phase of a group of pixels that can be connected in phase by using the phase image and the selected mask image as inputs and outputting a connected phase image.

In a tenth aspect of the present invention, in the above-mentioned eighth or ninth aspect, the phase image and the connection phase are output to a plurality of phase-connectable continuous regions having a definite phase value by using the output of the phase connection means as an input. An image is input, one continuous region is selected, and a phase value group of the continuous region is set as a globally defined phase value group. One continuous area closest to the area is selected, and a linear function is applied to the phase value groups for which the phase value groups of both continuous areas have not been determined while appropriately adding or subtracting an integer multiple of 2π, and the residual sum of squares is minimized. Is determined as an integral multiple of 2π radians, and is added to or subtracted from the undetermined phase value group to obtain a globally defined phase value group, and the above processing is repeated until all continuous regions take the globally defined phase value group. With a wide-area deterministic phase image And output.

According to an eleventh aspect of the present invention, in any one of the seventh to ninth aspects of the invention, after the deterministic phase value output from the phase connecting means is input and all the group of pixels that can be phase-connected are determined as the deterministic phase value, A phase image and a definite phase image are input, and a definite phase value in a group of pixels determined to be difficult to phase-connect is linearly estimated from the definite phase value around the pixel, and the estimated phase value and the phase connection are determined to be difficult. Phase determining means for adjusting the integer multiple of 2π radians so as to minimize the phase difference between the pixels, and repeating the above operation until all of the pixel groups determined to be difficult to phase-connect have a determined phase value and outputting as a determined phase image. Further included.

According to a twelfth aspect of the present invention, in the tenth aspect of the present invention, the phase image and the globally determined phase image are input after the globally determined phase value is input and all the group of pixels that can be connected to the phase are determined as the globally determined phase value. As an input, a globally determined phase value in a group of pixels determined to be difficult to phase connect is linearly estimated from the globally determined phase value around the pixel, and the phase difference between the estimated phase value and the pixel determined to be difficult to phase connect is most likely. The image processing apparatus further includes phase determination means for increasing or decreasing the integral multiple of 2π radians so as to reduce the number of pixels, and repeating the above operation until all of the pixel groups determined to be difficult to phase-connect have a globally determined phase value, and outputting as a globally determined phase image.

[0037]

Embodiments of the present invention will be described.

First, the phase connection method according to the first invention comprises the following steps.

Step 1: The phase image obtained by the phase shift method is input, a test area of a predetermined size including the pixel of interest is set in the phase image, and the phase of the pixel of interest and the pixels in the test area are set. If all the phase differences are smaller than the predetermined threshold value, it is determined that the target pixel can be connected in phase.

Step 2: It is determined whether or not the above-described phase connection possibility determination processing has been performed on all the pixels. If there is an unprocessed pixel, the process proceeds to step 1.

Step 3: One pixel is selected from a group of pixels that can be connected in phase, and the phase value of the pixel is set as a definite phase value.

Step 4: The phase is not determined with respect to the other group of pixels that can be connected to a phase, and a pixel whose adjacent pixel has a determined phase value is selected, and the other pixels that can be connected to a phase are connected to the determined phase. An integer multiple of 2π radian is added or subtracted so that the difference from the value becomes smaller than 2π radian, and the determined phase value is obtained.

Step 5: It is determined whether or not all the phase connectable pixel groups have a definite phase value. If any pixel group does not have a definite phase value, the process returns to step 4 and the process is repeated.

The processing in each of these steps can be realized by control of a program executed on a computer, and the FD or CD-ROM on which the program is recorded is stored.
The present invention also includes a recording medium such as a ROM. In this case, the present invention can be implemented by a computer reading and executing the program from the recording medium.

The first aspect of the present invention is characterized in that the phases can be connected at high speed and safely by previously selecting only a group of pixels to which the phases can be connected without inconsistency among the obtained phases.

In FIG. 4, reference numeral 106 denotes a partial phase image which is a part of the phase image. At this time, it is assumed that the range from black to white is assigned from -π radian to π radian. A test area having a predetermined size is set centering on each pixel on the image, and the connection possibility is determined.

For example, in the area A107 in FIG.
07 If the difference between the phase of the central pixel and the phase of the other pixels in the area A is smaller than the threshold value λ, it is determined that the central pixel of the area A107 is connected in phase. However, at this time, it should be noted that the phase is a periodic value that takes only a value from -π radian to π radian.

The phase of the pixel of interest is α ₀ , and the phase of the pixel in the area A 107 is α _i (i∈C _A ).

For the phase α _i , the following equation (7) or (8)
Is true, it is determined that the difference between the phases α ₀ and α _i is within the threshold λ.

| Α _i -α ₀ | <π ∩ | α _i -α ₀ | <λ (7)

| Α _i -α ₀ ± 2π | <π ∩ | α _i -α ₀ ± 2π | <λ (8)

If the phase difference of all the pixels in the area A107 is within the threshold value λ, it is defined that the pixel of interest is determined to be phase connectable.

By this determination, for example, the region B108 in FIG.
As described above, even when the phase value is located at the boundary between -π radians and π radians, it is possible to make an appropriate determination.

Here, the difference between the phase of the target pixel and the phases of all other pixels in the test area is within the threshold value. May be used as the threshold processing for the value expressing

Although the description has been made assuming that the target pixel is located at the center position of the test area, the target pixel need not be located at the center but may be included in the test area.

Further, in FIG. 4, the shape of the test area is rectangular, but another shape such as a circle or an ellipse may be used.

Of course, not all pixels in the test area but the phases of a specified number of pixels may be subjected to threshold processing.

The first pixel having a fixed phase value is:
It is selected from a group of pixels that can be connected in phase, but the selection method may be to scan the image and take the first pixel that can be connected in phase, or to give the pixel position from the outside. You may do so.

Next, the phase connection method according to the second invention comprises the following steps.

Step 1: The phase image obtained by the phase shift method is input, a test area of a predetermined size including the pixel of interest is set in the phase image, and the phase of the pixel of interest and the pixels in the test area are set. If all the phase differences are smaller than the predetermined threshold value, it is determined that the target pixel can be connected in phase.

Step 2: Control is performed so that the above-described phase connection possibility determination processing is performed on all pixels. That is,
It is determined whether or not the phase connection possibility determination processing has been performed on all the pixels. If there is any unprocessed pixel, the process proceeds to step 1.

Step 3: When a group of pixels that can be connected in phase forms a plurality of continuous regions, one continuous pixel group is selected first.

Step 4: One pixel is selected from the selected group of phase-connectable pixels, and the phase value of the pixel is set as a definite phase value.

Step 5: The phase is not determined for the other selected group of pixels that can be connected in phase, and an adjacent pixel selects a pixel having a determined phase value, and the other pixels can be connected in phase. An integral multiple of 2π radian is added or subtracted so that the difference between the pixel and the deterministic phase value is smaller than 2π radian, thereby obtaining a deterministic phase value.

Step 6: Steps 4 to 5 are repeated until all of the selected continuous pixel groups have a definite phase value.

Step 7: The processing of the above steps 3 to 6 is repeated until all the phase-connectable pixel groups have a definite phase value.

The processing in each of these steps can be realized by control of a program executed on a computer, and the present invention includes a recording medium on which the program is recorded. In this case, the present invention can be implemented by a computer reading and executing the program from the recording medium.

In the second aspect of the invention, it is implicitly assumed that the phase-connectable pixel groups all constitute continuous areas adjacent to each other. The second invention is characterized in that the phase is determined for each continuous area even if the continuous area is divided into a plurality of continuous areas.

In the second invention, as in the first invention, the method of selecting the pixel having the deterministic phase value first for each continuous area may be any method as long as it is a method of selecting the pixel.

Next, the phase connection method according to the third invention comprises the following steps.

Step 1: The phase image obtained by the phase shift method is input, a test area of a predetermined size including the pixel of interest is set in the phase image, and the phase of the pixel of interest and the pixels of the test area are set. If all the phase differences are smaller than the predetermined threshold value, it is determined that the target pixel can be connected in phase. Step 2: Control is performed so that the above-described phase connection possibility determination processing is performed on all pixels. Step 3: When a group of pixels that can be connected in phase forms a plurality of continuous regions, one continuous pixel group is selected from one or a plurality of continuous pixel groups specified. Step 4: One pixel is selected from the selected group of phase-connectable pixels, and the phase value of the pixel is set as a definite phase value. Step 5: The phase is not determined with respect to the other selected group of phase-connectable pixels, and an adjacent pixel selects a pixel having a determined phase value to determine the other phase-connectable pixel. An integer multiple of 2π radians is added or subtracted so that the difference from the phase value is smaller than 2π radians, and the determined phase value is obtained. Step 6: Steps 4 and 5 are repeated until all of the selected continuous pixel groups have a definite phase value. Step 7: Steps 3 to 6 are repeated until one or a plurality of designated phase-connectable pixels have a definite phase value.

The processing in each of these steps can be realized by control of a program executed on a computer, and the present invention includes a recording medium on which the program is recorded. In this case, the present invention can be implemented by a computer reading and executing the program from the recording medium.

The third invention is different from the second invention described above.
Rather than mechanically determining a definite phase value for all continuous regions, only a specified continuous region is processed.

As a method of designating a continuous area, for example, a small continuous area that is determined to be connectable due to noise or the like by threshold processing on the shape of each continuous area, such as the area and the vertical and horizontal sizes, is used. Any method may be used as long as it is a method of selecting a continuous area according to the purpose, for example, by selecting a continuous area having the largest area and selecting only an object to be measured which is largely photographed on the screen.

Next, the fourth invention comprises the following steps.

Step 11: The deterministic phase value output by the phase connection method according to the second or third aspect of the invention is input, and a plurality of phase-connectable continuous regions having the deterministic phase value are
One continuous area is selected, and a phase value group of the continuous area is set as a wide-area fixed phase value group. Step 12: One continuous area which is closest to the continuous area having the globally determined phase value group is selected from the continuous areas having no globally determined phase value group. Step 13: An integer multiple of 2π radian is added to or subtracted from the phase value group of the continuous region having no wide-area fixed phase value group so that the phase value group of both continuous regions is connected most smoothly. Step 14: The above steps 12 to 14 are repeated until all the continuous regions take the group of wide-area determinable phase values.

According to the second and third aspects of the present invention, it is only necessary to independently determine a definite phase value group for each of a plurality of continuous regions formed by a phase-connectable pixel group. Remained unknown.

According to the fourth aspect, the phase value groups of the respective continuous areas are smoothly connected to each other. The continuous region that does not have the globally defined phase value group and is closest to the continuous region that has the globally defined phase value group can be determined by the distance between the positions of the center of gravity of the continuous region.

Further, by combining a wide-area determined phase value group and a phase value group obtained by adding or subtracting an integral multiple of 2π radians to a phase value group that is not the wide-area determined phase value group, for example, by applying a linear function or the like, the residual square sum By defining an offset that is an integral multiple of 2π radians that is small, it is possible to smoothly connect to each other.

FIG. 5 is a diagram for explaining a method for determining a wide-area determinate phase value in the fourth invention. Referring to FIG. 5, a wide-area fixed phase value A111 and a fixed phase value A / a wide-area fixed phase candidate value A11 for which a relative phase value has not yet been determined.
Suppose you have two. Here, wide-area fixed phase candidate values B to D113 to 115 are generated by adding or subtracting an integer multiple of 2π radians to the fixed phase value A / wide-area fixed phase candidate value A.

Approximate straight lines A to D116 to 119 are obtained by applying a linear function by combining the wide-area fixed phase value A111 and one of the wide-area fixed phase candidate values A to D112 to 115.

The sum of squares of the difference between the approximate straight line A116, the wide-area determinable phase value A111, and the wide-area determinate phase candidate value A112 is defined as the residual sum of squares, which serves as a criterion for selecting the wide-area determinate phase candidate value A112. Similarly, the residual sum of squares is obtained for the other global fixed phase candidate values B to D113 to 115, and the candidate value having the smallest residual square sum (the global fixed phase candidate value C114 in FIG. 5) is set to the global fixed phase value. Is determined.

Although the above description has been made in a one-dimensional manner for simplicity, even a phase value that spreads two-dimensionally on a phase screen can be easily realized simply by replacing a straight line with a plane.

In the fourth invention, one continuous area is selected first. However, for example, a continuous area having the largest area may be selected or a selected continuous area may be provided from the outside. The selection method is not limited as long as the selection method is used.

Further, the continuous area is selected based on the distance between the continuous areas. However, the method of selecting one continuous area for determining the smoothness is not limited. It is needless to say that a plurality of continuous areas can be easily extended to the processing target.

As a method of determining a smooth connection,
Even in a method other than the above, for example, the phase value is extrapolated from the wide-area determinable phase value group to the outside of the region by a linear function, and the difference between the extrapolated phase value and the phase value group in another continuous region is the smallest. An integral multiple of 2π radians may be determined so as to be smaller.

In the above method, the function is a linear function. However, the function is not limited to the linear function, but may be any other smooth function.

Next, the fifth invention comprises the following steps.

Step 21: The deterministic phase value output by the phase connection method according to the first to third aspects of the present invention is input, and the deterministic phase value of a pixel group determined to be difficult to phase-connect is estimated from the deterministic phase value. Step 22: Add or subtract an integer multiple of 2π radian so that the phase difference between the estimated phase value and the pixel determined to be difficult to phase-connect is minimized. Step 23: The process of the above step 22 is repeated until all the pixel groups determined to be difficult to phase-connect have a definite phase value.

The fifth invention is characterized in that the phase value of a pixel group determined to be difficult to phase-connect is also determined.

FIG. 6 shows a fifth embodiment of the present invention.
FIG. 4 is a diagram for explaining a phase estimation method. Referring to FIG. 6, it is assumed that in the partial phase image 109, it is determined that the phase connection of the region C110 is difficult. However, estimation can be performed based on the determined phase value in the vicinity.

Next, a method of estimating and determining the phase in the embodiment of the present invention will be described with reference to FIG.
In FIG. 7, the wide-area determinate phase value A111 and the wide-area determinate phase value B1
It is assumed that 20 is a wide-area fixed phase value of a pixel group determined to be phase connectable. It is determined that the phase connection is difficult between the two wide-area determinable phase values, and the phase value in this region is determined by two undetermined phase values such as A122, B123, C124, and D125.
It is assumed that it is not determined which of the values is obtained by adding or subtracting an integral multiple of π radians.

The wide-area determinate phase value A111 and the wide-area determinate phase value B120
Between the estimated phase value 121 and the estimated phase value 121, and the phase value in the phase connection difficult region is determined by the undetermined phase value C having the smallest difference from the estimated phase value 121.
Assuming that it is 124, it is set as the final phase value.

Although the description has been made in a one-dimensional manner for simplicity, even a phase value that spreads two-dimensionally on a phase screen can be easily realized simply by replacing a straight line with a plane.

In the fifth invention, the estimation method is straight line estimation. However, other interpolation and extrapolation methods such as quadratic and cubic are also used.
Applicable.

The sixth invention comprises the following steps.

Step 31: In the fourth invention, the globally determined phase value output by the phase connecting method is input, and the globally determined phase value of a pixel group determined to be difficult to phase-connect is estimated from the globally determined phase value. Step 32: Add or subtract an integer multiple of 2π radians so that the phase difference between the estimated phase value and the pixel determined to be difficult to phase-connect is minimized. Step 33: The process of step 32 is repeated until all of the pixel groups determined to be difficult to phase-connect have a globally determined phase value.

In the fourth to sixth inventions, the processing of each step can be realized by control of a program executed on a computer, and the present invention includes a recording medium on which the program is recorded. In this case, each invention can be implemented by a computer reading and executing the program from the recording medium.

A phase connection apparatus according to a seventh aspect of the present invention receives a phase image calculated by a phase shift method, a test area size signal, and a threshold signal, and performs a test of a predetermined size including a target pixel in the phase image. Set the region, if the difference between the phase of the pixel of interest and the phase of the pixel in the test area are all smaller than a predetermined threshold, it is determined that the pixel of interest can be connected to the phase,
A phase connection test unit (11 in FIG. 1) that performs the phase connection possibility determination process on all the pixels and outputs a mask image composed of the phase connectable pixels, and receives the phase image and the mask image as inputs. A pixel is selected from a group of phase-connectable pixels, and the phase value of the pixel is set as a definite phase value, and the phase is not determined from the group of phase-connectable pixels, and a neighboring pixel has a definite phase value. Is selected, and an integer multiple of 2π radians is added or subtracted so that the difference between the undetermined phase value and the determined phase value becomes smaller than 2π radians to obtain a determined phase value. Phase connection means (12 in FIG. 1) for repeatedly outputting a connection phase image until the connection phase image has a value.

In the seventh aspect, the difference between the phase of the target pixel and the phases of all other pixels in the test area is within the threshold value. What is necessary is just a threshold process. Instead of all the pixels in the test area, the phases of a specified number of pixels may be set as the threshold processing target. Although the pixel having the final phase value at the beginning is selected from the group of pixels that can be connected in phase, the selection method may be such that the image is scanned and the pixel that can be connected first is found. Alternatively, the pixel position may be given from outside.

According to an eighth aspect of the present invention, in a case where a group of pixels capable of phase connection constitutes a plurality of continuous areas instead of the phase connection means in the phase connection device of the seventh invention, a phase image and a mask image are input and one A continuous pixel group is selected, one of the pixels in the continuous area is selected, and the phase value of the pixel is set as a definite phase value. Is selected, and an integer multiple of 2π radians is added or subtracted so that the difference between the undetermined phase value and the determined phase value is smaller than 2π radians to obtain a determined phase value. A phase connection unit that repeats the above operation until all continuous regions have a definite phase value, and outputs a connected phase image.

In the eighth invention as well, as in the seventh invention, the method of selecting a pixel whose definite phase value is first determined for each continuous region is not particularly limited as long as it is a method of selecting a pixel.

A ninth aspect of the present invention is directed to a phase connection device according to the eighth aspect of the present invention, in which, when a plurality of continuous areas are formed by a group of pixels that can be connected in phase, a mask image is input and one Continuous area selecting means for selecting one and outputting a selected mask image; and phase connecting means for outputting a connected phase image by obtaining a fixed value of the phase of a group of pixels that can be connected in phase with the phase image and the selected mask image as inputs. Prepare.

As a method of designating a continuous area in the ninth aspect, for example, a continuous area having a maximum area or a continuous area is selected in accordance with the purpose, such as threshold processing for the shape of each continuous area such as the area and the size of the horizontal and vertical directions. Any method is acceptable.

According to a tenth aspect, a phase image and a connection phase image are input to a plurality of phase-connectable continuous regions having a definite phase value, using the output of the phase connection device according to the eighth or ninth aspect as an input. And selects one continuous region and sets the phase value group of the continuous region as the globally defined phase value group, and is the closest to the continuous region having the globally defined phase value group among continuous regions having no globally defined phase value group. One continuous area is selected, and a linear function is applied to the phase value groups for which the phase value groups of both continuous areas are not determined, while appropriately adding or subtracting an integer multiple of 2π, and the residual square sum of 2π radians is the smallest. A wide-area fixed phase value group is obtained by adding or subtracting the undetermined phase value group by obtaining an integer multiple to obtain a wide-area fixed phase value group, and repeating the above operation until all continuous regions take the wide-area fixed phase value group. Output.

As a method of selecting a continuous region in the tenth invention, any method may be used as long as one continuous region is selected, for example, a continuous region having a maximum area or a selected continuous region is externally provided. Although the continuous area is selected based on the distance between the continuous areas, any method may be used as long as one continuous area for determining the smoothness is selected. Of course, it is easy to extend a plurality of continuous areas to be processed.

The method for determining the smooth connection is not limited to the above method. For example, the phase value may be shifted from the wide-area fixed phase value group to outside the area.
It is a matter of course that an extrapolation may be made by a linear function, and an integral multiple of 2π radian may be determined so that the difference between the extrapolated phase value and the group of phase values in the other continuous areas is minimized. Although the function is a linear function in the above method, the function is not limited to a linear function as long as it is a smooth function.

According to an eleventh aspect of the present invention, a phase image and a final phase image are obtained by inputting the final phase value output from the phase connection apparatus of the seventh to ninth aspects and setting all the phase-connectable pixels to the final phase value. Is input, the deterministic phase value in the group of pixels determined to be difficult to phase connect is linearly estimated from the determinate phase value around the pixel, and the estimated phase value and the phase difference between the pixels determined to be difficult to phase connect are the smallest. The phase determining means (14 in FIG. 8) which adds and subtracts an integral multiple of 2π radians and repeats the above operation until all the pixel groups determined to be difficult to phase-connect have a determined phase value and outputs a determined phase image. Prepare.

According to a twelfth aspect of the present invention, a phase image and a wide-area determinable phase image are obtained by setting the output of the phase connection device according to the sixth to ninth aspects as an input, and setting all the phase-connectable pixels to a global deterministic phase value. As an input, the estimated phase value in the group of pixels determined to be difficult to phase-connect is linearly estimated from the wide-area deterministic phase value around the pixel, and the estimated phase value and the phase difference between the pixels determined to be difficult to phase-connect are the smallest. The phase determining means (16 in FIG. 9) which adds and subtracts an integral multiple of 2π radians and repeats the above operation until all the pixel groups determined to be difficult to phase-connect have a determined phase value and outputs as a determined phase image. Prepare.

In the eleventh invention, the estimation method is straight line estimation, but another interpolation method such as quadratic or cubic may be used.

According to a twelfth aspect of the present invention, a phase image and a wide-area determined phase image are obtained after the globally determined phase value output from the phase connection device of the tenth aspect is input and all the group of pixels that can be phase-connected are defined as the globally determined phase value. Is input, a globally determined phase value in a group of pixels determined to be difficult to phase connect is linearly estimated from the globally determined phase value around the pixel, and the phase difference between the estimated phase value and the pixel determined to be difficult to phase connect is A phase determining means for adding or subtracting an integer multiple of 2π radians so as to be the smallest, repeating the above operation until all the pixel groups determined to be difficult to phase-connect have a globally determined phase value, and outputting as a globally determined phase image (FIG. 9). 16) is provided. Hereinafter, the present invention will be described in detail with reference to examples.

[0112]

FIG. 1 is a block diagram showing the configuration of a first embodiment of the phase connection device of the present invention. It is assumed that the phase value has already been calculated based on the above equation (6), and a phase image 105 as shown in FIG.

The phase connection test means 11 outputs the phase image signal
1 and the test area size signal 4 and the phase threshold signal 5 as inputs, a test is performed to determine whether or not the phase difference in the test area is within the threshold based on the above equations (7) and (8). Run
A connectable pixel group is output as a connectable mask image signal 2.

The phase connection means 12 receives the phase image signal 1 and the connectable mask image signal 2 as inputs, scans the connectable mask image signal 2, and determines the phase value of the first connectable pixel found as a fixed value. The phase value of adjacent connectable pixels is adjusted by an integer multiple of 2π radian so that the difference between the phase value and the final phase value is 2π radian or less.

The above operation is repeated until all the connectable pixels have the determined phase value, and then output as the connection phase image signal 3.

In the first embodiment, the test as to whether or not the connection is possible is performed based on the equations (7) and (8). Anything can constitute the present invention. Similarly, the present invention can be configured such that all pixels in the region are set, but the phase values of a specified number of pixels are set as threshold value processing targets. The method of selecting a pixel having the final determined phase value first may be any method as long as one pixel is determined, and the present invention can also be configured by providing a pixel position from the outside.

In the first embodiment, it is implicitly assumed that the number of continuous regions formed by connectable pixel groups appearing on the connectable mask image signal 2 is one. When a plurality of continuous areas appear, the connectable mask image signal 2 is not scanned by the phase connecting means 12 in the first embodiment to obtain a definite value, and the connectable pixel found first is connected. Is a definite value, and the phase value of an adjacent connectable pixel is defined as
Add or subtract an integer multiple of 2π radians so that it is less than or equal to π radians. If all connectable pixels that form a continuous area have a definite phase value, repeat the above operation for a continuous area that does not have another definite value When all the phase values have become the definite values, it is possible to respond by outputting the connection phase image signal 3.

Similarly, when there are a plurality of continuous regions, the phase connection processing is not performed for all the continuous regions. For example, it is determined that a region having a small area is accidentally determined to be connectable due to noise or the like. The connection processing may be performed only on the area equal to or larger than the threshold value, or only the continuous area having the largest area may be selected with the object photographed largest on the image as the processing target.

The selection processing of the region to be subjected to the phase connection processing may include threshold processing for the length in the vertical and horizontal directions, and the like.
What is necessary is just a continuous area selection process according to the purpose.

FIG. 8 is a block diagram showing the configuration of a second embodiment of the phase connection device of the present invention. As in the first embodiment, the phase value has already been calculated based on the above equation (6), and the phase image 105 as shown in FIG. I do.

The phase connection test means 1 outputs the phase image signal 1
And the test area size signal 4 and the phase threshold signal 5 as inputs, a test is performed based on the above equations (7) and (8) to determine whether the phase difference in the test area is within the threshold. Then, a connectable pixel group is output as a connectable mask image signal 2.

The phase connection means 12 receives the phase image signal 1 and the connectable mask image signal 2 as inputs, scans the connectable mask image signal 2 and does not take a definite value, and detects the first connectable mask image signal. The phase value of a pixel is defined as a definite value, and the phase values of adjacent connectable pixels are adjusted by adding or subtracting an integral multiple of 2π radians so that the difference from the definite phase value is 2π radians or less. If all the pixels have a definite phase value, the above operation is repeated for a region that does not have another definite value, and if all the phase values have become definite values, they are output as connected phase image signals 3.

The wide-area phase determining means 14 receives the connection phase image signal 3 as input, selects a continuous area having the largest area among the continuous areas appearing in the signal 3, and converts the phase value group of the continuous area into a wide area. A group of deterministic phase values. In addition, the center of gravity of each continuous region is obtained, and a continuous region closest to the continuous region having the globally determined phase value group is selected from the continuous regions having no globally determined phase value group. An integer multiple of 2π is appropriately added to or subtracted from the group of phase values that are not globally determined, and a linear function is applied together with the group of phase values that are globally determined.

At this time, an integral multiple of 2π radian at which the residual sum of squares of the fitted linear function and the phase value group to be fitted is minimized is calculated, and the phase value group that has not been globally determined is added to or subtracted from the phase value group. Let it be a phase value. This operation is repeated until all the continuous regions have the wide-area determinate phase value.
It is output as the wide-area fixed phase signal 6.

In the second embodiment of the present invention, the test as to whether or not connection is possible was performed based on the above equations (7) and (8).
Other than that, threshold processing may be performed on a value expressing a phase variation such as a standard deviation, and is not limited to the above equations (7) and (8). Similarly, all pixels in the region are set, but the phase values of a specified number of pixels may be set as the threshold value processing target. The method of selecting a pixel having a final phase value first is not limited as long as one pixel is determined, and a pixel position may be externally given.

In the second embodiment, the continuous area having the largest area is selected first. However, any method may be used to select one continuous area, for example, by giving a selected continuous area from the outside. Although the continuous area is selected based on the distance between the continuous areas, a method of selecting one continuous area for determining smoothness may be used, and the selection method itself is not particularly limited. It is also easy to extend a plurality of continuous areas to be processed.

FIG. 9 is a block diagram showing the configuration of a third embodiment of the phase connection device of the present invention. As in the above embodiment, the phase value has already been calculated based on the above equation (6),
The phase image 105 as shown in FIG.
Is input.

The phase connection test means 11 receives the phase image signal 1, the test area size signal 4 and the phase threshold signal 5 as inputs, and performs a test as to whether or not the phase difference in the test area is within the threshold value according to the above equation. This is executed based on (7) and the above equation (8), and a connectable pixel group is output as a connectable mask image signal 2. The phase connection means 12 receives the phase image signal 1 and the connectable mask image signal 2 as inputs, and
Is not taken as a definite value, and the phase value of the connectable pixel found first is set as the definite value, and the phase value of the adjacent connectable pixel is set to a difference of 2π radians or less from the definite phase value. If all consecutive connectable pixels have a definite phase value, repeat the above operation for a region that does not have another definite value, and then add and subtract integer multiples of 2π radians so that When the value becomes a definite value, it is output as the connection phase image signal 3.

The wide-area phase determining means 14 receives the connection phase image signal 3 as input, selects a continuous area having the largest area among the continuous areas appearing in the signal, and broadly determines the phase value group of the continuous area. A group of phase values. In addition, the center of gravity of each continuous region is obtained, and a continuous region closest to the continuous region having the globally determined phase value group is selected from the continuous regions having no globally determined phase value group. 2 for phase values not globally determined
An integer multiple of π is appropriately adjusted, and a linear function is applied together with the group of wide-area deterministic phase values. At this time, an integer multiple of 2π radian that minimizes the residual sum of squares of the fitted linear function and the phase value group to be fitted is obtained, and is added to or subtracted from the phase value group that is not globally determined to be a globally determined phase value. . This operation is repeated until all the continuous regions have the globally determined phase value, and are output as the globally determined phase signal 6.

The phase determining means 16 receives the phase image signal 1 and the wide-area determined phase signal 6 as inputs, and linearly estimates a determined phase value in a group of pixels determined to be difficult to phase-connect from the wide-area determined phase value around the pixel. Then, 2π is set so that the phase difference between the estimated phase value and the pixel determined to be difficult to connect to the phase is minimized.
An integer multiple of radians is added or subtracted, and the above operation is repeated until all the pixel groups determined to be difficult to phase-connect have a deterministic phase value, and output as a deterministic phase image signal 7.

In the third embodiment, the test as to whether or not connection is possible was performed based on the above equations (7) and (8). In addition, the variation of the phase such as the standard deviation is expressed. Any threshold processing may be applied to the value to be processed. Similarly, all the pixels in the region are set, but the phase value of a specified number of pixels may be set as a threshold value processing target. In addition, the method of selecting a pixel having a final phase value at the very beginning may be a method of determining one pixel, and for example, a pixel position may be given from the outside.

In the third embodiment, the continuous area having the largest area is selected first. However, a method of selecting one continuous area, such as giving a selected continuous area from the outside, may be used. . Although the continuous area is selected based on the distance between the continuous areas, any method may be used as long as one continuous area for determining the smoothness is selected. It is also easy to extend a plurality of continuous areas to be processed.

In the third embodiment, the phase determining means 16 is added to the second embodiment.
A configuration may be adopted in which the phase determining means 16 is added to the embodiment.

[0134]

As described above, according to the present invention,
The connection relation of the phases calculated by the phase shift method can be determined accurately and at high speed, and an effect that a more practical three-dimensional shape measuring apparatus can be realized is achieved.

The reason is that, in the present invention, the phase connection test means inputs the phase image, the size of the test area, and the threshold value, and determines whether or not the phase difference in the test area is within the threshold value. Judgment is made based on (7) and the above equation (8), a connectable pixel group is output as a connectable mask image signal, and the phase connection means receives the phase image and the connectable mask image as inputs, and And the phase value of the first connectable pixel found by scanning is defined as a definite value, and the phase values of adjacent connectable pixels are set to 2π radians such that the difference from the definite phase value is 2π radians or less. This is because the above operation is repeated until all connectable pixels have the determined phase value, and then output as a connected phase image.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a first exemplary embodiment of the present invention.

FIG. 2 is a schematic diagram of a three-dimensional shape measuring apparatus based on a grid pattern method.

FIG. 3 is an example of a phase image obtained by calculation.

FIG. 4 is a diagram for explaining a phase connection test method according to the present invention.

FIG. 5 is a diagram for explaining a method for determining a phase of a plurality of continuous regions according to the present invention.

FIG. 6 is a diagram for explaining a phase estimation method according to the present invention.

FIG. 7 is a diagram for explaining a phase estimation method according to the present invention.

FIG. 8 is a block diagram showing a configuration of a second exemplary embodiment of the present invention.

FIG. 9 is a block diagram showing a configuration of a third exemplary embodiment of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 phase image signal 2 connectable mask image signal 3 connected phase image signal 4 test area size signal 5 phase threshold signal 6 wide area fixed phase image signal 7 fixed phase image signal 10 phase connection device 11 phase connection test means 12 phase connection means 13 Phase connection device 14 Wide-area phase determination means 15 Phase connection device 16 Phase determination means 100 Object 101 Light source 102 Lattice 103 Camera 104 Point P 105 Phase image 106 Partial phase image 107 Area A 108 Area B 109 Partial phase image 110 Area C 111 Wide area determination Phase value A 112 Determined phase value A / Wide-area fixed phase candidate value A 113 Wide-area fixed phase candidate value B 114 Wide-area fixed phase candidate value C 115 Wide-area fixed phase candidate value D 116 Approximate line A 117 Approximate line B 118 Approximate line C 119 Approximate Straight line D 120 Wide-area determinate phase value B 121 Estimated phase value 1 22 Undetermined phase value A 123 Undetermined phase value B 124 Undetermined phase value C 125 Undetermined phase value D

Claims (19)

(57) [Claims] (A) A phase image obtained by a phase shift method is input, a test area of a predetermined size including a pixel of interest is set in the phase image, and the phase of the pixel of interest and the inside of the test area are set. When all of the differences from the phase of the pixel are smaller than a predetermined threshold value, determining that the target pixel can be connected in phase; and (b) for all the pixels of the phase image , Said (a)
(C) selecting one pixel from the group of phase-connectable pixels obtained in the steps (b) and (b), and setting the phase value of the pixel to a definite phase value. And (d) selecting a pixel whose phase is not determined for a group of pixels other than the one pixel that can be connected in phase and whose adjacent pixels have a determined phase value, a step of a definite phase value difference of the possible pixel and the definite phase value by adding or subtracting an integer multiple of 2π radians to be less than 2π radians, the process steps of (e) above (d), the phase connection Repeating until all possible pixel groups have a definite phase value. 2. A phase image obtained by a phase shift method is input, a test area of a predetermined size including a target pixel is set in the phase image, and the phase of the target pixel and the test area in the test area are set. When all of the pixel phase differences are smaller than a predetermined threshold value, determining that the target pixel can be connected in phase; and (b) determining (P) a) performing a phase connection possibility determination process; and (c) when the group of phase connectable pixels acquired in the steps up to (b) constitutes a plurality of continuous regions, one of the Selecting a continuous pixel group; (d) selecting one pixel from the selected phase-connectable pixel group and setting the phase value of the pixel to a definite phase value; Other than the pixel, the selected phase-connectable pixel A pixel whose phase is not determined for the group and whose adjacent pixels have a determined phase value is selected, and 2π is selected so that the difference between the other phase-connectable pixels and the determined phase value is smaller than 2π radian. (F) repeating the processing of the step (e) until all the selected continuous pixel groups have a deterministic phase value, and (g) Repeating the processing of the steps (c) to (f) until all the phase-connectable pixel groups have a definite phase value. 3. A phase image obtained by the phase shift method is input, a test area of a predetermined size including a target pixel is set in the phase image, and the phase of the target pixel and the test area in the test area are set. If all of the pixel phase differences are smaller than a predetermined threshold, determining that the pixel of interest is phase connectable; and (b) determining (a) for all pixels in the phase image. And (c) determining one continuous pixel group from one or more continuous pixel groups designated when the group of phase-connectable pixels forms a plurality of continuous regions. (D) selecting one pixel from the selected group of phase-connectable pixels and setting the phase value of the pixel to a definite phase value; and (e) other than the one pixel. For the selected phase-connectable pixel group of Is not determined, and a pixel whose adjacent pixel has a determined phase value is selected, and an integer multiple of 2π radian is added or subtracted so that the difference between the other phase-connectable pixels and the determined phase value is smaller than 2π radian. (F) repeating the processes of the above (d) to (e) until all the selected continuous pixel groups have a deterministic phase value; (C) repeating the steps (c) to (f) until one or a plurality of designated phase-connectable pixels have a definite phase value. 4. The phase connection method according to claim 2, wherein (h) inputting the deterministic phase value and forming one continuation region for a plurality of phase-connectable continuation regions having the deterministic phase value. Selecting and setting a phase value group of the selected continuous region to a global fixed phase value group; and (i) a continuous region having a global fixed phase value group among continuous regions not having the global fixed phase value group. And (j) selecting 2 radians for the phase value group of the continuous region that does not have the globally defined phase value group so that the phase value group of both continuous regions is connected most smoothly. And (k) repeating the above steps (h) to (j) until all continuous regions take a group of globally defined phase values. . 5. The phase connection method according to claim 1, wherein (1) inputting the deterministic phase value, and determining the deterministic phase value in a pixel group determined to be difficult to phase-connect to the deterministic phase value. (M) adding or subtracting an integer multiple of 2π radian so that the phase difference between the estimated phase value and the pixel determined to be difficult to phase-connect is minimized; Repeating the step (m) until all the pixel groups determined to be difficult to phase-connect have a definite phase value. 6. The phase connection method according to claim 4, wherein (o) inputting the globally determined phase value and estimating the globally determined phase value of a pixel group determined to be difficult to phase-connect from the globally determined phase value. (P) adding and subtracting an integer multiple of 2π radian so that the phase difference between the estimated phase value and the pixel determined to be difficult to phase-connect is minimized; and (q) the (o) and (p). And b.) Repeating the step of) until all the pixel groups determined to be difficult to phase-connect have a globally determined phase value. 7. A phase image calculated by the phase shift method, a test area size signal, and a threshold signal are input, and a test area of a predetermined size including a pixel of interest in the phase image is set. When the difference between the phase of the pixel of interest and the phase of the pixel in the test area is all smaller than the predetermined threshold value, the phase of the pixel of interest is determined to be phase connectable. A phase connection test unit that performs the phase connection possibility determination process on all pixels in the phase image and outputs a mask image composed of the phase connectable pixels; and inputs the phase image and the mask image. Then, one pixel is selected from a group of phase-connectable pixels, and the phase value of the pixel is set as a definite phase value, and a pixel whose phase is not determined and a neighboring pixel has a definite phase value from the group of phase-connectable pixels Select An integer multiple of 2π radians is added or subtracted so that the difference between the undetermined phase value and the determined phase value becomes smaller than 2π radians to obtain a determined phase value. And a phase connecting means for acquiring and outputting a connected phase image by repeating the above steps. 8. A phase image calculated by a phase shift method, a test area size signal, and a threshold signal are input, and a test area of a predetermined size including a target pixel is set in the phase image. If the difference between the phase of the pixel of interest and the phase of the pixel in the test area is all smaller than a predetermined threshold, a phase connection determination process is performed to determine that the pixel of interest can be phase-connected, and all of the phase images A phase connection test unit that performs the phase connection possibility determination process on the pixel of (a) and outputs a mask image composed of the phase connectable pixels; and a case where the group of phase connectable pixels forms a plurality of continuous regions. The phase image and the mask image are input, one continuous pixel group is selected, one pixel in the continuous region is selected, and the phase value of the pixel is set as a definite phase value, and the phase value is selected from the continuous pixel group. Is not fixed A pixel whose adjacent pixel has a definite phase value is selected, and the difference between the undetermined phase value and the definite phase value is 2
An integer multiple of 2π radians is added / subtracted to be smaller than π radians to obtain a final phase value, and the above process is repeated until all the continuous pixel groups have final phase values. Phase connection means for acquiring and outputting a connected phase image by repeating the processing. 9. The selected mask image according to claim 8, wherein when the group of pixels that can be phase-connected constitute a plurality of continuous areas, a mask image is input and one of the continuous areas is selected. Continuous region selecting means for outputting the phase image and the selection mask image, and a phase connecting means for obtaining a fixed value of the phase of a group of pixels that can be phase connected and outputting a connected phase image. A phase connection device characterized by the following. 10. The phase connection device according to claim 8, wherein an output of said phase connection means is input, and a phase image and a connection phase image are applied to a plurality of phase-connectable continuous regions having a definite phase value. Is input, one continuous region is selected, and a phase value group of the continuous region is set as a global fixed phase value group, and a continuous region having a global fixed phase value group among continuous regions having no global fixed phase value group Is selected, and a linear function is applied to the phase value groups for which the phase value groups of both continuous regions have not been determined, while appropriately adjusting the integer multiple of 2π, so that the residual sum of squares is minimized. By obtaining an integer multiple of 2π radians and adding and subtracting the undetermined phase value group to obtain a globally determined phase value group, the above process is repeated until all continuous regions take the globally determined phase value group. Wide-area deterministic phase image Output to the phase connection and wherein the. 11. The phase connection device according to claim 7, wherein the deterministic phase value output from the phase connection means is used as an input.
After setting all the phase-connectable pixels to the deterministic phase value, the phase image and the deterministic phase image are input, and the deterministic phase value of the pixel group determined to be difficult to phase-connect is linearly determined from the deterministic phase value around the pixel. Estimation is performed such that the phase difference between the estimated phase value and the pixel determined to be difficult to connect to the phase is minimized.
Phase connection further comprising phase determination means for adding or subtracting an integral multiple of π radians and repeating the above operation until all the pixel groups determined to be difficult to phase-connect have a determined phase value and outputting as a determined phase image. apparatus. 12. The phase connection device according to claim 10, wherein the globally determined phase value is input, and a group of pixels that can be phase-connected are all determined as a globally determined phase value, and then a phase image and a globally determined phase image are input. The linearly estimated phase value in the group of pixels determined to be difficult to phase connection is linearly estimated from the globally determined phase value around the pixel, and the phase difference between the estimated phase value and the pixel determined to be phase connection difficult is the smallest. 2π to be
A phase determining means for adding or subtracting an integer multiple of radians, repeating the above operation until all the pixels determined to be difficult to phase-connect have a globally determined phase value, and outputting as a globally determined phase image. Phase connection device. 13. Determining a connection relation of phases calculated by a phase shift method using a data processing device, wherein: (a) a phase image signal calculated by a phase shift method;
When a signal for determining the size of the test area and a threshold signal are input, and the difference between the phase of the pixel of interest in the phase image and the phase of the pixel in the test area is all smaller than a predetermined threshold It is determined that the target pixel can be connected in phase,
A connectable pixel group is output as a connectable mask image signal. (B) The phase image signal and the connectable mask image signal are input, and the connectable mask image signal is scanned to find the connectable first. The phase value of a pixel that can be connected is determined to be a definite value, and the phase value of an adjacent connectable pixel
a process of adding / subtracting an integral multiple of 2π radians so as to be smaller than or equal to π radians, and repeatedly outputting the connected phase image signal until all connectable pixels have a definite phase value. 14. A process executed when determining a connection relation of phases calculated by the phase shift method in the three-dimensional shape measurement: (a) inputting a phase image obtained by the phase shift method; A test area of a predetermined size including the target pixel is set, and the difference between the phase of the target pixel and the phase of the pixel in the test area is all smaller than a predetermined threshold. A process of determining that the pixel of interest is phase connectable; (b) a process of performing the phase connectability determination process of (a) on all pixels of the phase image; a process of selecting one pixel from the group of phase-connectable pixels obtained in the steps up to b) and setting the phase value of the pixel to a definite phase value; and (d) other phases other than the one pixel. Phase for connectable pixel group Confirmed Izu, the selected pixels adjacent pixel has definite phase value, the other by addition or subtraction of an integral multiple of 2π radians so that the difference between the phase connection possible pixel definite phase value is less than 2π radians And (e) controlling to repeat the process (d) until all of the group of pixels that can be phase-connected have the determined phase value. A) a recording medium on which a program for causing a computer to execute each process of the above) is recorded. 15. A process executed when determining a connection relation of phases calculated by a phase shift method in three-dimensional shape measurement, includes the steps of: (a) inputting a phase image obtained by a phase shift method; A test area of a predetermined size including the target pixel is set, and when all of the difference between the phase of the target pixel and the phase of the pixel in the test area are smaller than a predetermined threshold value, A process of determining that a pixel can be connected in phase; (b) a process of controlling to perform the phase connection possible determination process of (a) on all pixels of the phase image; and (c) a process of (b). And (d) selecting one continuous pixel group from among the plurality of continuous-connection-capable pixel groups obtained from the above-described processing when the plurality of continuous-pixel groups form a plurality of continuous regions. One pixel from pixel group (E) selecting a phase value of the pixel and setting the phase value of the pixel to a determined phase value; and (e) adjacent to each other other than the one pixel, the phase of which is not determined for the selected group of connectable pixels. Select a pixel whose matching pixel has a definite phase value, and add or subtract an integer multiple of 2π radian so that the difference between the other phase-connectable pixels and the definite phase value is smaller than 2π radian. (G) a process of controlling the process of (e) to be repeated until all of the selected continuous pixel groups have a definite phase value; and (g) a process of (c) to (f). And a recording medium for storing a program for causing a computer to execute the above-described processes (a) to (g) until all of the phase-connectable pixel groups have a determined phase value. 16. A process executed when determining a connection relation of phases calculated by a phase shift method in three-dimensional shape measurement, wherein (a) a phase image obtained by a phase shift method is input;
A test area of a predetermined size including the target pixel in the phase image is set, and if all the differences between the phase of the target pixel and the phases of the pixels in the test area are smaller than a predetermined threshold, A process of determining that a pixel can be connected in phase; (b) a process of controlling to perform the phase connection possible determination process of (a) on all the pixels of the phase image; A process of selecting one continuous pixel group from one or a plurality of specified continuous pixel groups when the pixel group constitutes a plurality of continuous regions; and (d) from the selected phase-connectable pixel group. (E) a process of selecting one pixel and setting the phase value of the pixel to a determined phase value; and (e) determining the phase of the selected other group of pixels that can be connected to the phase other than the one pixel. Pixels whose adjacent pixels have a definite phase value (D) adding and subtracting an integral multiple of 2π radians to obtain a final phase value so that the difference between the other phase-connectable pixels and the final phase value is smaller than 2π radians; ) To (e) are controlled so as to repeat until the selected continuous pixel group has all deterministic phase values; and (g) the processes (c) to (f) are designated. A recording medium on which is recorded a program for causing a computer to perform the processing of controlling to repeat until one or a plurality of phase-connectable pixel groups have a definite phase value, and the processing of (a) to (g). 17. The recording medium according to claim 15, wherein (h) inputting the deterministic phase value and selecting one continuous area from a plurality of phase-connectable continuous areas having the deterministic phase value. And processing the phase value group of the selected continuous region to be a globally defined phase value group; and (i) converting a continuous region having a globally defined phase value group to a continuous region having no globally defined phase value group. (J) a process of selecting one closest continuous region; and (j) a phase value group of 2π radians for a phase value group of a continuous region having no wide-area fixed phase value group so that the phase value groups of both continuous regions are connected most smoothly. (H) through (h) through (h) through (j) controlling to repeat the above processes (h) through (j) until all continuous regions take a group of wide-area determinable phase values. The process of k) is executed by a computer. Recording medium on which a program for recording is further recorded. 18. The recording medium according to claim 14, wherein (l) inputting the deterministic phase value, and determining the deterministic phase value of a pixel group determined to be difficult to connect to the phase from the deterministic phase value. (M) a process of adding or subtracting an integral multiple of 2π radians so that the phase difference between the estimated phase value and a pixel determined to be difficult to phase-connect is minimized; and (n) the above (l) and (l). a) a process for controlling the process of m) to be repeated until all of the pixel groups determined to be difficult to phase-connect have a definite phase value; and a program for causing the computer to execute the above processes (l) to (n). Further recorded recording medium. 19. The recording medium according to claim 17, wherein: (o) inputting the globally determined phase value, and estimating the globally determined phase value of a pixel group determined to be difficult to phase-connect from the globally determined phase value. (P) a process of adding or subtracting an integral multiple of 2π radians so that the phase difference between the estimated phase value and the pixel determined to be difficult to phase-connect is minimized; and (q) the (o) and (p). A recording medium further recording a program for causing a computer to execute the above processes (o) to (q) until all the pixel groups determined to be difficult to phase-connect have a globally determined phase value; .