JP5968370B2 - Three-dimensional measuring apparatus, three-dimensional measuring method, and program - Google Patents

Description

  The present invention relates to a three-dimensional measurement apparatus, a three-dimensional measurement method, and a three-dimensional measurement method for measuring a three-dimensional shape by a spatial encoding method that obtains distance information by projecting a plurality of pattern lights in which bright and dark portions are arranged with arbitrary widths, and Regarding the program.

  There is known a pattern projection method in which a plurality of pattern lights are projected onto a subject, and the subject on which the pattern light is projected is imaged to perform shape measurement using the principle of triangulation from pattern deformation. Among them, the spatial encoding method that projects the stripe pattern light in which the bright part and the dark part are alternately arranged with an arbitrary width and binary encodes the space is often used in the field of three-dimensional measurement, and commercialized. It has also been made.

  With reference to FIG. 1, the structure of a general three-dimensional shape measuring apparatus and the concept of three-dimensional measurement will be described. The three-dimensional shape measuring apparatus generally includes a projector 101 that irradiates a measurement target object with a pattern and a camera 102 that images a reflection pattern. Stripe pattern light in which bright portions and dark portions are alternately arranged with an arbitrary width is projected from the projector 101 toward the subject 103. The fringe pattern light has a plurality of predetermined pattern shapes, which are projected and captured by the camera 102 each time and acquired as image data. The boundary position between the bright part and the dark part of the subject 103 is assumed to be (X, Y, Z). Assume that the principal point position of the projector 101 when the boundary position (X, Y, Z) and the projector 101 are connected by a line is (X1, Y1). Similarly, the principal point position of the camera 102 when the boundary position (X, Y, Z) and the camera 102 are connected by a line is defined as (X2, Y2). Here, the principal point position (X2, Y2) of the camera 102 is a point obtained from the horizontal coordinates of the image sensor (CCD, CMOS, etc.) of the camera 102. The horizontal coordinate is determined by the horizontal width and the vertical width of the image sensor. For example, in the case of a 640 × 480 image sensor, the horizontal x coordinate has a value of 0 to 640, and the vertical y coordinate has a value of 0 to 480. Similarly, the principal point position (X1, Y1) of the projector 101 is also obtained from the horizontal coordinates of the light projection sensor. The distance L between the projector 101 and the camera 102 is a base line length and is a value determined from the configuration conditions of the apparatus. From these parameters, the boundary position (X, Y, Z) of the subject 103 can be obtained by the principle of triangulation. By obtaining the respective boundary positions (X, Y, Z) with respect to the entire surface of the subject 103, the three-dimensional shape of the subject 103 can be measured.

  Next, the shape of the stripe pattern light will be described. With reference to FIG. 2, the stripe pattern light of a binary code called a Gray code having encoding error resistance will be described. A portion observed in black in the captured reflection pattern corresponds to 0, and a portion observed in white corresponds to 1. In FIG. 2A, the whole is divided into two and encoded as 1 and 0 in two areas. In FIG. 2B, the four areas of the bright part and the dark part are encoded as 1, 0, 0, 1, and the corresponding stripe pattern light is irradiated and imaged. Further, in FIG. 2C, eight regions are encoded as 1, 0, 0, 1, 1, 0, 0, 1 and the corresponding stripe pattern light is irradiated and imaged. Thus, since the encoded area number is assigned to each area, each area can be determined. Each area is (1,1,1), (1,1,0), (1,0,0), (1,0,1), (0,0,1), (0,0,0) , (0, 1, 0), (0, 1, 1). Since the space can be divided into eight by using the three stripe pattern lights in this way, this spatial coding is referred to as 3-bit spatial coding. The stripe pattern lights in FIGS. 2A, 2B, and 2C are referred to as 1-bit stripe pattern light, 2-bit stripe pattern light, and 3-bit stripe pattern light, respectively.

  When measuring the three-dimensional shape in more detail, n stripe pattern lights are irradiated while sequentially reducing the bright and dark areas. Each area can be determined by assigning an area number obtained by dividing the irradiation area of the projector into 2n. For example, 10-bit spatial encoding is performed in three-dimensional measurement in which a region is divided into 1024 regions.

  The advantage of the Gray code is that even if a coding error occurs at the boundary of the region due to a shift or blur of the fringe pattern light, it is judged as an adjacent region and no coding error occurs as a greatly shifted region. is there. For this reason, spatial coding using the Gray code is a commonly used method.

  In order to improve accuracy in the three-dimensional measurement by the spatial encoding method, it is necessary to determine the horizontal coordinate position x (hereinafter referred to as “boundary position”) of the boundary between the bright part and the dark part from the captured image data. With reference to FIG. 3, the boundary position between the bright part and the dark part will be described. FIG. 3 is a graph when the luminance of image data obtained by imaging 2-bit stripe pattern light is plotted on the vertical axis and the horizontal coordinate x is plotted on the horizontal axis. FIG. 3A shows an ideal boundary position. In FIG. 3A, since the brightness values of the bright part and the dark part are different by 1 and 0 at the boundary, the boundary position between the bright part and the dark part is uniquely determined, and the position a and the position b can be determined as the boundary position. However, in actual measurement, the vicinity of the boundary changes gently as shown in FIG. 3B due to the blur of the stripe pattern light, the reflectance of the subject, the influence of external light, etc., and the boundary position can be uniquely determined. This is not possible and causes measurement errors.

  In order to avoid such a problem, a method is disclosed in which a reliability indicating measurement accuracy is given to a measurement point, and the measurement coordinate is not used if the reliability is equal to or less than a threshold (for example, see Patent Document 1). .

  In Patent Document 1, in addition to the stripe pattern light, the subject is irradiated with all lighting light that is the entire bright portion and all extinction light that is the entire dark portion. The reliability is defined as the ratio of the brightness difference of the stripe pattern light to the brightness difference between the all lighting light and all the lighting light. Only measurement points whose reliability exceeds a threshold are output.

Japanese Patent No. 3996560

  In the invention of Patent Document 1, it is necessary to obtain the luminance difference of the stripe pattern light. In order to accurately obtain the luminance difference, it is necessary to calculate a luminance curve using the luminance values of a plurality of pixels, and there is a problem that the calculation cost is high. There is also a method of obtaining a luminance difference by determining sampling points in advance, but in that case, the reliability is different for each sampling point, and there is a problem that the reliability may not be accurately reflected.

  In view of the above problems, an object of the present invention is to easily and accurately calculate a reliability indicating measurement accuracy.

The three-dimensional measuring apparatus according to the present invention that achieves the above object is as follows:
A first image in which the bright part region and the dark areas are arranged alternately pattern is obtained by imaging the target object is projected, the pattern and the bright area and the dark area of the pattern is reversed is projected Image acquisition means for acquiring a second image obtained by imaging the target object ;
Determine the boundary position between the bright region and the dark region of the pattern the a picture pixel value that is part of the first images, and based on the the image pixel value to the Ru contained in the second images determining means for,
And image pixel value that is part of the first images, and based on the the image pixel value to the Ru contained in the second images, the reliability acquisition for acquiring a reliability indicating the accuracy of the boundary position Means,
Deriving means for deriving three-dimensional shape information of the target object based on the reliability.

The three-dimensional measuring apparatus according to the present invention that achieves the above object is as follows:
A first image obtained by imaging a target object on which a pattern in which bright areas and dark areas are alternately arranged is projected, and the target object on which an entire illumination pattern including only a bright area is projected. a second image obtained by capturing an image acquisition means for acquiring a third image obtained by imaging the target object in a state where the pattern is not projected,
Wherein the first images, and the second images Contact and said bright area and the dark region of the pattern in said first image which is determined based on the third images A determining means for determining the boundary position ;
And based on the pixel values of at least two pixels of the neighborhood of the boundary position of the first image, a reliability obtaining means for obtaining a reliability indicating the accuracy of the boundary position,
Deriving means for deriving three-dimensional shape information of the target object based on the reliability.

  According to the present invention, since the reliability indicating the measurement accuracy is obtained by using a plurality of pixels near the position where the brightness is inverted from the acquired image information, the reliability can be easily and accurately calculated. Become. In addition, since the reliability can be obtained simultaneously with the calculation of the boundary position, the calculation process can be speeded up.

The conceptual diagram of the structure and measuring method of a three-dimensional shape measuring apparatus. The figure which shows the fringe pattern light which the space encoding method by a Gray code | symbol projects. Explanatory drawing of a boundary position. FIG. The figure which shows the reliability calculation which concerns on 1st Embodiment. The figure which shows the reliability calculation which concerns on 2nd Embodiment. The figure which shows the reliability calculation which concerns on 3rd Embodiment.

(First embodiment)
With reference to FIG. 4, the basic composition schematic of the three-dimensional shape measuring apparatus which concerns on this invention is demonstrated. The three-dimensional measurement apparatus includes a projector 402 that projects a stripe pattern light in which bright portions and dark portions are alternately arranged onto a subject 407, and a camera that captures reflected pattern light of the subject 407 (target object) onto which the stripe pattern light is projected. 403 and a calculation processing unit 41 that executes various operations. The calculation processing unit 41 instructs the projection and imaging of the fringe pattern light, and calculates the captured image data.

  The calculation processing unit 41 includes a CPU 400, a pattern memory 401, an image memory 1404, a reliability calculation unit 405, and a measurement processing unit 406. The CPU 400 executes various calculations. The pattern memory 401 is composed of a ROM or the like, and stores a pattern shape program for stripe pattern light, a time account program for setting a projection time, and the like. The pattern memory 401 that has received the projection instruction from the CPU 400 transmits a stripe pattern light shape signal to the projector 402. In addition, a time account signal is transmitted to the projector 402 and the camera 403, and the timing of projection and imaging of the stripe pattern light is managed.

  Image data captured by the camera 403 is temporarily stored in the image memory 404 and sequentially transmitted to the reliability calculation unit 405. The reliability calculation unit 405 is configured by a ROM, a RAM, or the like, and performs a process of determining the boundary position between the bright part and the dark part of the stripe pattern light and calculating the reliability of the boundary position. A method for determining the boundary position and a method for calculating the reliability will be described later.

  Next, the image data is transmitted to the measurement processing unit 406. The measurement processing unit 406 calculates three-dimensional position information from the image data. The measurement processing unit 406 includes a ROM, a RAM, and the like. The measurement processing unit 406 stores parameters depending on devices such as the baseline length, focal length, and the number of pixels of the projector 402 and the camera 403, distortion by calibration performed in advance, external parameters such as external light luminance, and the like. It also stores a three-dimensional measurement program by triangulation. The measurement processing unit 406 binarizes the image data to generate code data. When performing spatial encoding with N bits, the stripe pattern light has N types of pattern shapes, and thus N pieces of code data are generated. The reliability and the boundary position calculated by the reliability calculation unit 405 are given to the data. Next, the correspondence relationship between the pixels of the projector 402 and the pixels of the camera 403 is taken from the code data, and three-dimensional measurement is performed by the principle of triangulation to obtain three-dimensional shape information (three-dimensional shape calculation). The measurement result is generated as distance image data corresponding to the number of pixels of the camera 403. The reliability can be arbitrarily given to the distance image data. The distance image data is imaged by an output unit (not shown) and can be observed as a distance image by a display unit (not shown).

  The processes of the image memory 404, the reliability calculation unit 405, and the measurement processing unit 406 are all performed based on an instruction from the CPU 400.

  Next, a method for calculating the reliability will be described with reference to FIG.

  FIG. 5A shows the stripe pattern light to be projected. The fringe pattern light 51 is a spatial pattern fringe pattern light using a 3-bit gray code, and is referred to as a positive pattern. The stripe pattern light 52 is reverse stripe pattern light obtained by inverting the bright and dark portions of the stripe pattern light 51 and is referred to as a negative pattern. In this figure, a positive / negative pattern of 3-bit stripe pattern light is shown as an example. However, in measurement by actual spatial coding, for example, a positive / negative pattern of stripe pattern light of 1 to 10 bits may be projected and photographed according to the resolution of the projector. That is, if the measurement is 10 bits, 20 shots are required. The concept of determining the boundary position will be described with reference to FIG. FIG. 5B is a graph in which the vertical axis represents the luminance of the captured image data and the horizontal axis represents the horizontal pixel coordinate. Image data obtained by imaging the reflection pattern light of the stripe pattern light is defined as first image data. Similarly, image data obtained by imaging the reflected pattern light of the reverse stripe pattern light is set as second image data. A luminance line 53 is a luminance line of a part of the first image data imaged with the stripe pattern light 51. On the other hand, the luminance line 54 is the luminance line of the second image data imaged with the reverse stripe pattern light 52, and the relative position of the horizontal pixel coordinates is the same as that of the luminance line 53. The luminance line 53 and the luminance line 54 intersect at a position M and a position N. Therefore, the position M and the position N are set as boundary positions. An actual boundary position calculation method will be described with reference to FIG. Pixel numbers are placed below the graph, and points 55 and 56 are pixels of image data captured by projecting the s-th and s + 1-th stripe pattern light 51, respectively. Points 57 and 58 are pixels of image data obtained by projecting the sth and s + 1th fringe pattern light 52, respectively. Then, the points 55 and 56 and the points 57 and 58 are connected by straight lines, respectively, and the horizontal coordinates of the intersection 59 of the line segment 500 and the line segment 501 are obtained. In this way, the boundary position N is obtained. Here, the reliability is defined as an angle φ formed by the intersection of the line segment 500 and the line segment 501.

  That is, the first luminance value for the fringe pattern light 51 and the second luminance value for the reverse fringe pattern light 52 are calculated for each pixel constituting each image data of the first image data and the second image data. (Luminance value calculation). Then, the position corresponding to the pixel whose magnitude relationship between the first luminance value and the second luminance value is inverted is determined as the boundary position between the bright part and the dark part. A first luminance gradient of the stripe pattern light 51 at the boundary position and a second luminance gradient of the reverse stripe pattern light 52 are calculated. Each luminance gradient calculation process may be a process of calculating the gradient of each luminance curve at the boundary position. A correlation value indicating the correlation between the first luminance gradient and the second luminance gradient is calculated as a reliability indicating the accuracy of the boundary position. Note that the correlation value (reliability) is not limited to the angle φ, and may be, for example, the degree of coincidence between the line segment 500 and the line segment 501 at the boundary position of the luminance curve in which each luminance value is plotted.

Hereinafter, the calculation method of the angle φ which is the reliability will be described in detail. The coordinates of each point are point 55 (Xa, Ya), point 56 (Xb, Yb), point 57 (Xa, Yd), and point 58 (Xb). , Yc). Then, an intersection of the line segment 500 and the line segment 501 is set as a point 59 (Px, Py).

  The slope P1 of the line segment 500 is

  The slope P2 of the line segment 501 is

It is obtained.
A formula of a straight line passing through the line segment 500 from the formulas (1) and (2) is expressed as a formula (3).

Similarly, the equation of a straight line passing through the line segment 501 is expressed as equation (3).

The coordinates (Px, Py) of the point 59 from the expressions (3) and (4) are expressed as the expressions (5) and (6).

  The boundary position is calculated by equation (5).

  When the angle φ is divided into φ1 and φ2 as shown in FIG. 5C, φ1 is expressed by the equation (7) and φ2 is expressed by the equation (8) from the coordinates of each point.

  From the equations (7) and (8), the angle φ is obtained as in the equation (9).

  Here, if the distance between the pixels is 1, Xb−Xa = 1, and therefore Expression (9) can be expressed as Expression (10).

  Thereby, the reliability φ (angle φ) is calculated.

  Based on the calculated reliability, the measurement processing unit 406 prohibits output on the distance image for measurement points whose reliability is a predetermined threshold or less. In addition, reliability is given to each measurement point of the distance image data and all points are output, and when the distance image data is edited, the measurer arbitrarily sets a threshold value of reliability, so that the desired high reliability measurement can be performed. It is also possible to acquire distance image data of only points.

  As described above, according to the first embodiment of the present invention, since the angle formed by the line segments obtained by linearly connecting the four pixels in the nearest vicinity of the inversion position of the light and dark area from the image data is set as the reliability, The reliability can be calculated at the same time as obtaining the boundary position. Further, since the calculation can be obtained only with the luminance values of the four points, it can be easily calculated.

  In the first embodiment, the angle between the line segments is used as the reliability. However, the luminance value of the intersection of the two line segments (point 59 in FIG. 5) may be added as the second reliability. . The luminance value of the point 59 can be obtained by Expression (6). If the reliability is only an angle, the brightness of the reflected light from the object cannot be known. Even at the same angle, the measurement accuracy differs between high brightness and low brightness, and at high brightness above the threshold, the contrast may decrease and reliability may decrease even at the same angle. Therefore, the luminance value at the intersection of the two line segments is also set as the reliability, and the determination accuracy based on the reliability can be further improved by determining it together with the angle.

(Second Embodiment)
In this embodiment, a reliability calculation method of a spatial encoding method using a positive pattern and a uniform pattern light with a uniform light amount will be described.

  FIG. 6A shows the stripe pattern light to be projected. The fringe pattern light 601 is spatial pattern fringe pattern light using a 3-bit gray code. A pattern 602 is all illumination pattern light of only the bright part. A pattern 603 is all-off pattern light only in the dark part. In the case of performing 3-bit spatial encoding, 1-bit to 3-bit stripe pattern light, all illumination patterns 602 only for bright portions, and all extinction patterns 603 only for dark portions are required. Image data obtained by imaging the reflection pattern light of the stripe pattern light is defined as first image data. Similarly, image data obtained by imaging the reflection pattern light of all illumination pattern light is set as second image data. Further, the image data obtained by imaging the reflection pattern light of the all-off pattern light is set as second image data.

  A method for determining the boundary position of the stripe pattern light 601 will be described with reference to FIG. FIG. 6B is a graph in which the luminance value of the captured image data is plotted on the vertical axis and the horizontal coordinate x is plotted on the horizontal axis.

  A luminance line 53 is a luminance line (first luminance value curve) of the first image data imaged with the stripe pattern light 601. A luminance line 604 is a luminance line (second luminance value curve) of second image data captured by projecting the entire illumination pattern 602. A luminance line 605 is a luminance line (third luminance value curve) of third image data captured by projecting the all-off pattern 603. An average luminance value of the luminance line 604 and the luminance line 605 is obtained and set as a luminance line 606. It is assumed that the luminance line 53 of the stripe pattern light 601 and the average luminance line 606 intersect at a position M ′ and a position N ′. Therefore, the position M ′ and the position N ′ are set as boundary positions.

Next, an actual calculation method will be described with reference to FIG. Pixel numbers are placed below the graph, and points 607 and 608 are pixels of image data captured by projecting the s-th and s + 1-th stripe pattern light 601 respectively. Points 609 and 610 are pixels of image data obtained by projecting the s-th and s + 1-th all illumination patterns 602, respectively. Points 611 and 612 are pixels of image data obtained by projecting the s-th and s + 1-th all-off patterns 603, respectively. When the average luminance value of the corresponding pixels of the image data captured by projecting the entire illumination pattern 602 and the all-off pattern 603 is obtained, the average value of the sth point 609 and the point 611 is the point 61
3 is obtained. Further, the average value of the s + 1-th point 610 and the point 612 is obtained as a point 614. Therefore, the points 607 and 608, the points 613 and 614 are connected by straight lines, and the horizontal coordinate of the intersection 615 is obtained. In this way, the boundary position N ′ can be calculated. Here, the reliability is defined as an angle θ formed by the intersection of two straight lines. Similar to the first embodiment, the angle θ can be obtained by Expression (9). The reliability may be a luminance gradient, that is, a gradient at the boundary position, instead of the angle θ.

  Also in the second embodiment, the luminance value of the intersection (point 615) of two line segments may be added as the second reliability.

  As described above, according to the second embodiment, reliability can be obtained in the spatial coding method using a positive pattern and a uniform pattern light having a uniform light amount as in the first embodiment. This embodiment is effective in measurement requiring a higher speed than the accuracy because the number of shots is about half, although the accuracy is reduced as compared with the spatial encoding method using a positive / negative pattern.

(Third embodiment)
In the third embodiment, a reliability calculation method when a pixel in the vicinity of the inversion position of the light and dark area is selected will be described.

  FIG. 7 is a graph in which the brightness of captured image data is taken on the vertical axis and the horizontal coordinate x is taken on the horizontal axis. A luminance line 53 is a luminance line of a part of image data picked up with a positive pattern. A luminance line 54 is a luminance line of image data captured with a negative pattern. The relative position of the horizontal pixel coordinates is the same as that of the luminance line 53. The line segment 500 and the line segment 501 are the line segments obtained by linearly coupling the luminances of the pixels closest to the inversion position of the light and dark area, that is, the pixel numbers s and s + 1. In this case, the intersection angle φ between the line segment 500 and the line segment 501 is defined as the reliability.

  In the above description, the pixel closest to the inversion position of the light / dark region is used. However, a pixel in which the pixel interval is doubled may be selected. Points 702 and 703 are pixels of image data obtained by projecting the s−1 and s + 2 positive patterns, respectively. Points 704 and 705 are pixels of image data obtained by projecting the s−1 and s + 2 negative patterns, respectively. The points 702 and 703 and the points 704 and 705 are connected by straight lines to form a line segment 706 and a line segment 707, respectively. In this case, the intersection angle φ ′ between the line segment 706 and the line segment 707 is set as the reliability.

  Furthermore, the pixel interval is further widened and tilted using pixels that are equidistant from the inversion position of the light-dark region, such as the s-2th pixel and the s + 3rd pixel, or the s-3th pixel and the s + 4th pixel. And the angle formed by the line segment may be similarly set as the reliability.

  As described above, according to the third embodiment, it is possible to select a plurality of pixels in the vicinity from the inversion position of the light and dark area. Therefore, a plurality of inclinations may be calculated as reliability, and the average value may be used as the reliability.

(Other embodiments)
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed.

Claims (20)

A first image obtained by imaging a target object onto which a pattern in which bright areas and dark areas are alternately arranged is projected, and a pattern in which the bright area and the dark area of the pattern are inverted is projected. Image acquisition means for acquiring a second image obtained by imaging the target object;
Determining means for determining a boundary position between the bright area and the dark area of the pattern based on a pixel value included in the first image and a pixel value included in the second image;
Reliability acquisition means for acquiring a reliability indicating the accuracy of the boundary position based on a pixel value included in the first image and a pixel value included in the second image;
Deriving means for deriving three-dimensional shape information of the target object based on the reliability. The reliability value obtaining means said among the pixel values of the pixels included in the first images, the pixel values of at least two pixels in the vicinity of the boundary position, a pixel of pixels included in the second images The three-dimensional measurement apparatus according to claim 1, wherein a reliability indicating the accuracy of the boundary position is calculated based on pixel values of at least two pixels in the vicinity of the boundary position among the values. The reliability acquisition means acquires a first line representing a change in pixel value based on pixel values of at least two pixels in the vicinity of the boundary position among the pixels included in the first image, A second line representing a change in pixel value is obtained based on pixel values of at least two pixels in the vicinity of the boundary position among the pixels included in the second image, and the first line and the first line The three-dimensional measurement apparatus according to claim 1, wherein the reliability is acquired based on an angle formed by a line of two.   The determination means determines a pixel position where a first line representing a change in pixel value included in the first image and a second line representing a change in pixel value included in the second image intersect. The three-dimensional measuring device according to claim 1, wherein the three-dimensional measuring device is determined as the boundary position. A first image obtained by imaging a target object on which a pattern in which bright areas and dark areas are alternately arranged is projected, and the target object on which an entire illumination pattern including only a bright area is projected. Image acquisition means for acquiring a second image obtained by imaging, and a third image obtained by imaging the target object in a state in which no pattern is projected;
A boundary position between the bright area and the dark area of the pattern in the first image determined based on the first image, the second image, and the third image is determined. A determination means;
Reliability acquisition means for acquiring reliability indicating the accuracy of the boundary position based on pixel values of at least two pixels in the vicinity of the boundary position of the first image;
Deriving means for deriving three-dimensional shape information of the target object based on the reliability. The determining means takes an average of a first line representing a change in a pixel value included in the first image, a pixel value included in the second image, and a pixel value included in the third image. The three-dimensional measurement apparatus according to claim 5, wherein a boundary position between the bright area and the dark area is determined based on a second line representing a change in average pixel value.   The three-dimensional measurement apparatus according to claim 6, wherein the reliability acquisition unit calculates the reliability based on an angle formed by the first line and the second line. A first image obtained by imaging a target object onto which a pattern in which bright areas and dark areas are alternately arranged is obtained is acquired, and a pattern in which the bright areas and dark areas of the pattern are inverted is obtained. Image acquisition means for acquiring a second image obtained by imaging the projected target object;
Based on a first line that approximates a change in pixel value in a first direction in the first image and a second line that approximates a change in pixel value in a second direction in the second image. Determining means for determining a position of an intersection of the first line and the second line;
Based on the pixel values of at least two pixels in the vicinity of the intersection among the pixels of the first image, and the pixel values of at least two pixels in the vicinity of the intersection of the pixels of the second image, A reliability acquisition means for acquiring a reliability indicating the accuracy of the position of the intersection of the first line and the second line to be determined;
A three-dimensional measurement apparatus comprising: a derivation unit that derives three-dimensional shape information of the target object based on the reliability.   The three-dimensional measurement apparatus according to claim 1, wherein the first image and the second image are sequentially captured by the same imaging unit. The derivation means obtains three-dimensional shape information for each position corresponding to the boundary position according to the principle of triangulation using the position of the projection means that projects the pattern and the position of the imaging means that images the target object. three-dimensional measuring apparatus according to any one of claims 1 to 7, wherein the deriving. It said deriving means, when the reliability is below the threshold for the boundary position, claims 1 to 7, further comprising a prohibiting means for prohibiting the output of the three-dimensional shape information of a position corresponding to the boundary position The three-dimensional measurement apparatus according to any one of the above. The three-dimensional measurement apparatus according to claim 11 , further comprising an acquisition unit configured to acquire the threshold value based on a user instruction. The three-dimensional measurement apparatus further includes a projecting unit that projects the pattern onto a subject.
Three-dimensional measuring apparatus according to any one of claims 1 to 12 and having an imaging means for imaging the object. The pixel value is three-dimensional measuring apparatus according to any one of claims 1 to 13, characterized in that a luminance value. A three-dimensional measurement method in a three-dimensional measurement apparatus,
A first image obtained by the image acquisition means imaging a target object on which a pattern in which bright areas and dark areas are alternately arranged is projected, and the bright area and the dark area of the pattern are inverted. An image acquisition step of acquiring a second image obtained by imaging the target object onto which the pattern is projected;
A determination that the reliability acquisition unit determines a boundary position between the bright area and the dark area of the pattern based on a pixel value included in the first image and a pixel value included in the second image; Process,
A reliability acquisition step of acquiring a reliability indicating the accuracy of the boundary position based on a pixel value included in the first image and a pixel value included in the second image;
And a deriving step of deriving three-dimensional shape information of the target object based on the reliability. A three-dimensional measurement method in a three-dimensional measurement apparatus,
The image acquisition means projects a first image obtained by imaging a target object onto which a pattern in which bright areas and dark areas are alternately arranged is projected, and an entire illumination pattern including only the bright areas. An image obtaining step of obtaining a second image obtained by imaging the target object and obtaining a third image obtained by imaging the target object in a state in which no pattern is projected;
A boundary between the bright area and the dark area of the pattern in the first image determined by the determining means based on the first image, the second image, and the third image A determining step for determining the position;
A reliability acquisition step in which a reliability acquisition means acquires reliability indicating the accuracy of the boundary position based on pixel values of at least two pixels in the vicinity of the boundary position of the first image;
And a deriving step of deriving three-dimensional shape information of the target object based on the reliability. A program for causing a computer to function as a three-dimensional measuring apparatus, the program
A first image obtained by imaging a target object onto which a pattern in which bright areas and dark areas are alternately arranged is projected, and a pattern in which the bright area and the dark area of the pattern are inverted is projected. Image acquisition means for acquiring a second image obtained by imaging the target object;
Determining means for determining a boundary position between the bright area and the dark area of the pattern based on a pixel value included in the first image and a pixel value included in the second image;
A reliability acquisition means for acquiring a reliability indicating the accuracy of the boundary position based on a pixel value included in the first image and a pixel value included in the second image;
A program that functions as derivation means for deriving three-dimensional shape information of the target object based on the reliability. A program for causing a computer to function as a three-dimensional measuring apparatus, the program
A first image obtained by imaging a target object on which a pattern in which bright areas and dark areas are alternately arranged is projected, and the target object on which an entire illumination pattern including only a bright area is projected. Image acquisition means for acquiring a second image obtained by imaging and a third image obtained by imaging the target object in a state in which no pattern is projected;
A boundary position between the bright area and the dark area of the pattern in the first image determined based on the first image, the second image, and the third image is determined. Decision means,
Reliability acquisition means for acquiring a reliability indicating the accuracy of the boundary position based on pixel values of at least two pixels in the vicinity of the boundary position of the first image;
A program that functions as derivation means for deriving three-dimensional shape information of the target object based on the reliability. A three-dimensional measurement method in a three-dimensional measurement apparatus,
An image acquisition means acquires a first image obtained by imaging a target object onto which a pattern in which bright area and dark area are alternately arranged is projected, and the bright area and dark area of the pattern An image acquisition step of acquiring a second image obtained by imaging the target object onto which the inverted pattern is projected;
A first line approximating a change in a pixel value in a first direction in the first image; and a second line approximating a change in a pixel value in a second direction in the second image. And determining a position of an intersection of the first line and the second line based on
The reliability acquisition means includes pixel values of at least two pixels in the vicinity of the intersection among the pixels of the first image, and pixel values of at least two pixels in the vicinity of the intersection of the pixels of the second image. And a reliability acquisition step of acquiring a reliability indicating the accuracy of the position of the intersection of the first line and the second line determined based on:
And a deriving step of deriving three-dimensional shape information of the target object based on the reliability. A program for causing a computer to function as a three-dimensional measuring apparatus, the program
A first image obtained by imaging a target object onto which a pattern in which bright areas and dark areas are alternately arranged is obtained is acquired, and a pattern in which the bright areas and dark areas of the pattern are inverted is obtained. Image acquisition means for acquiring a second image obtained by imaging the projected target object;
Based on a first line that approximates a change in pixel value in a first direction in the first image and a second line that approximates a change in pixel value in a second direction in the second image. Determining means for determining the position of the intersection of the first line and the second line;
Based on the pixel values of at least two pixels in the vicinity of the intersection among the pixels of the first image, and the pixel values of at least two pixels in the vicinity of the intersection of the pixels of the second image, A reliability acquisition means for acquiring a reliability indicating the accuracy of the position of the intersection of the first line and the second line to be determined;
A program that functions as derivation means for deriving three-dimensional shape information of the target object based on the reliability.

Images