JP6029376B2 - Precision measurement method - Google Patents

Description

  The present invention relates to a precision measurement method that uses a camera to accurately measure between lattice points on a plane that is an object to be measured.

  In recent years, in order to measure the shrinkage rate of an object such as concrete, there is a demand for grasping not only the distance between points but also surface displacement information. By performing surface management, it is possible to know the deformation state of the object more accurately. On the other hand, the resolution of digital input devices represented by digital cameras has become very high.

  A method of calibrating a camera at the time of image measurement using a digital camera is disclosed (for example, refer to Patent Documents 1 and 2). An image deformation processing method is also known (see Non-Patent Document 1, for example). Furthermore, measurement of concrete shrinkage using a digital camera has also been attempted (see, for example, Non-Patent Documents 2 and 3).

JP 2003-307466 A Japanese Patent Laid-Open No. 9-329418

Author Masahito Masatatsu "All Image Processing Algorithms Implemented in C Language [Details] Image Processing Programming" Publishing SoftBank Creative Masakazu Uchino et al. “Development of full-field measurement method for structure diagnosis using optical deformation measurement method” Research Report No.16, 2006, Fukuoka Industrial Technology Center Yusuke Aoki “Fundamental Study for Measuring Shrinkage Strain Distribution on the Concrete Surface by Digital Image Shooting” Proceedings of Concrete Engineering Vol.31, No.1, 2009

  However, for example, in the methods for measuring concrete shrinkage using digital cameras in Non-Patent Documents 2 and 3, pattern matching is performed from images before and after shrinkage, and the displacement is examined. In this method, it is necessary to fix the camera before and after contraction, and the degree of freedom is low. In addition, since the positions before and after shrinkage are measured by pattern matching, measurement may not be possible depending on the state of the concrete surface (in the case of a region having no surface characteristics).

  SUMMARY OF THE INVENTION An object of the present invention is to provide a precision measuring apparatus and a precision measuring method that can eliminate the above-mentioned drawbacks and measure the deformation state of an object to be measured with high accuracy using a digital input device.

The precision measuring device according to the first aspect of the present invention includes:
(A) For the four vertices designated in a quadrangular shape around the object to be measured, the length between the four vertices is at least five of the four sides and two diagonal sides of the quadrangle. A measuring unit to obtain,
(B) Photographing the object to be measured and its surroundings with a camera so that the four vertices around the object to be measured and a plurality of grid points designated on the object to be measured are photographed. An imaging unit for obtaining images;
(C) a pre-correction image coordinate acquisition unit for obtaining image coordinates of the four vertices before tilt correction from the obtained captured image;
(D) a corrected image coordinate calculation unit for obtaining image coordinates of four vertices after tilt correction based on the length between the vertices;
(E) Projection transformation coefficient calculation unit for obtaining a projection transformation coefficient for tilt correction based on a relational expression between the image coordinates of four vertices before tilt correction and the image coordinates of the four vertices after tilt correction When,
(F) In a captured image, a grid point coordinate calculation unit before correction for obtaining image coordinates of the plurality of grid points on the object to be measured;
(G) A corrected lattice point coordinate calculation unit for obtaining image coordinates of the plurality of lattice points after tilt correction by performing projective conversion of the image coordinates of the plurality of lattice points using the obtained projective transformation coefficient. When,
including.

  In the precision measuring apparatus according to the present invention, it is only necessary to specify four vertices in a rectangular shape around the object to be measured, and since it is not limited to rectangular vertices, precise measurement with a high degree of freedom can be performed.

(A) is a block diagram which shows the structure of the precision measuring device which concerns on Embodiment 1, (b) is a block diagram which shows the physical structure of the control part of (a). FIG. 5 is a plan view showing four vertices designated in a square shape around the object to be measured and a plurality of lattice points designated on the object to be measured. It is a modification of FIG. 2, and is a plan view when four vertices constitute a quadrilateral other than a rectangle. It is the top view which extracted the square which consists of four vertices around the to-be-measured object. (A)-(j) is an example of a marker used as four vertices or a plurality of grid points. It is a figure which shows an example of the image which image | photographed four vertices. 3 is a flowchart of a precision measurement method according to the first embodiment. 3 is a flowchart of a computer program for precision measurement according to the first embodiment. 5 is a flowchart of a displacement measuring method according to Embodiment 2. 7 is a flowchart of a displacement measurement computer program according to the second embodiment.

The precision measuring device according to the first aspect of the present invention includes:
(A) For the four vertices designated in a quadrangular shape around the object to be measured, the length between the four vertices is at least five of the four sides and two diagonal sides of the quadrangle. A measuring unit to obtain,
(B) Photographing the object to be measured and its surroundings with a camera so that the four vertices around the object to be measured and a plurality of grid points designated on the object to be measured are photographed. An imaging unit for obtaining images;
(C) a pre-correction image coordinate acquisition unit for obtaining image coordinates of the four vertices before tilt correction from the obtained captured image;
(D) a corrected image coordinate calculation unit for obtaining image coordinates of four vertices after tilt correction based on the length between the vertices;
(E) Projection transformation coefficient calculation unit for obtaining a projection transformation coefficient for tilt correction based on a relational expression between the image coordinates of four vertices before tilt correction and the image coordinates of the four vertices after tilt correction When,
(F) In a captured image, a grid point coordinate calculation unit before correction for obtaining image coordinates of the plurality of grid points on the object to be measured;
(G) A corrected lattice point coordinate calculation unit for obtaining image coordinates of the plurality of lattice points after tilt correction by performing projective conversion of the image coordinates of the plurality of lattice points using the obtained projective transformation coefficient. When,
including.

The precision measurement method according to the second aspect of the present invention includes:
(A) A step of specifying four vertices in a square shape around the periphery 2 of the object to be measured 1 and measuring at least five sides of the four sides and two diagonal sides of the square as the length between the four vertices When,
(B) The object to be measured by the camera so that the four vertices 3 specified in a square shape around the object 2 and the plurality of lattice points 4 specified on the object 1 can be seen. Photographing the periphery to obtain a photographed image;
(C) obtaining image coordinates of four vertices before tilt correction from the obtained captured image;
(D) obtaining image coordinates of the four vertices after tilt correction based on the measured length between the vertices;
(E) Relationship between the image coordinates of the four vertices before the tilt correction obtained in step (c) and the image coordinates of the four vertices after the tilt correction obtained in step (d) Obtaining a projective transformation coefficient for tilt correction based on the equation;
(F) obtaining the image coordinates of the plurality of grid points on the object to be measured in the captured image;
(G) obtaining the image coordinates of the plurality of lattice points after tilt correction by performing projective transformation of the image coordinates of the plurality of lattice points using the obtained projective transformation coefficient;
including.

  The tilt correction method according to the third aspect is the step of (a) in the second aspect, wherein the length between the vertices in the step (a) is four sides and one diagonal side of the quadrangle, or three sides of the quadrangle. At least one of the side and the two diagonal sides is measured.

The computer program for precision measurement according to the fourth aspect of the present invention provides:
(A) The object to be measured and its surroundings so that the four vertices 3 specified in a square shape on the periphery 2 of the object to be measured 1 and the plurality of lattice points specified on the object to be measured 1 are shown. Obtaining image coordinates of four vertices before tilt correction from a photographed image obtained by photographing
(B) As the length between the four vertices, the image coordinates of the four vertices after tilt correction are calculated based on the length measured for at least five of the four sides and two diagonal sides of the rectangle. Seeking steps,
(C) Relationship between the image coordinates of the four vertices before the tilt correction obtained in step (a) and the image coordinates of the four vertices after the tilt correction obtained in step (b) Obtaining a projective transformation coefficient for tilt correction based on the equation;
(D) Obtain image coordinates of the plurality of grid points on the object to be measured in the photographed image. (E) Projective transform the image coordinates of the plurality of grid points using the obtained projective transformation coefficients. Thus, obtaining image coordinates of the plurality of grid points after tilt correction;
Let the computer execute each of these steps to perform precision measurement.

  A computer-readable recording medium according to the fifth aspect stores the computer program for precision measurement according to the fourth aspect.

In the displacement measurement method according to the sixth aspect, in the first time, the precision measurement method of the second aspect is performed, and images of the plurality of lattice points on the object to be measured in the first time are obtained. Determining the coordinates;
At a second time after the first time, the precision measurement method of the second aspect is performed to obtain image coordinates of the plurality of grid points on the measurement object at the second time. Steps,
Displacement over time from the image coordinates of the plurality of lattice points on the object to be measured at the first time to the image coordinates of the plurality of lattice points on the object to be measured at the second time A step of seeking
including.

A displacement measuring method according to a seventh aspect is the sixth aspect, wherein the camera is fixed from the first time to the second time,
In the step of obtaining image coordinates of the plurality of lattice points on the object to be measured at the second time, only steps (b), (f), and (g) in the precision measurement method according to claim 2. To obtain image coordinates of the plurality of grid points on the object to be measured.

  A computer program for displacement measurement according to an eighth aspect causes a computer to execute each step of the displacement measurement method according to the sixth aspect or the seventh aspect, and the measurement target is from the first time to the second time. The displacement with time of the image coordinates of a plurality of grid points on the object is measured.

  A computer-readable recording medium according to a ninth aspect stores the displacement measurement computer program according to the eighth aspect.

(Embodiment 1)
<Precision measuring device>
FIG. 1A is a block diagram showing the configuration of the precision measuring apparatus 10 according to the first embodiment. FIG. 1B is a block diagram illustrating an example of a physical configuration of the control unit 13. The precision measuring apparatus 10 photographs the measuring unit 11 that measures the length between four vertices 3 (A1 to A4) designated on the periphery 2 of the object 1 and the object 1 and its periphery 2. The photographing unit 12 that obtains the photographed image, the pre-correction image coordinate acquisition unit 14 that obtains the image coordinates of the four vertices before the tilt correction, and the post-correction image coordinate calculation unit that obtains the image coordinates of the four vertices after the tilt correction 15, a projective transformation coefficient calculation unit 16 for obtaining a projection transformation coefficient for tilt correction, a pre-correction grid point coordinate calculation unit 17 for obtaining image coordinates of a plurality of grid points 4 (B11 to Bnm) before tilt correction, And a corrected lattice point coordinate calculation unit 18 for obtaining image coordinates of a plurality of corrected lattice points. The pre-correction image coordinate acquisition unit 14, the post-correction image coordinate calculation unit 15, the projective transformation coefficient calculation unit 16, the pre-correction lattice point coordinate calculation unit 17, and the post-correction lattice point coordinate calculation unit 18 are: It may be realized as a function of the control unit 13. As shown in FIG. 1B, the control unit 13 includes a CPU 21, a memory 22, a storage unit 23, an input / output unit 24, a display unit 25, and a communication unit 26 as physical configurations.

  Below, each structural member of this precision measuring device is demonstrated.

  For the four vertices 3 (A1 to A4) designated in a quadrangular shape around the object 2 to be measured 1, the measuring unit 11 has a length between the four vertices, four sides of the quadrangle, and two diagonal sides. , Get a length of at least 5 sides. The imaging unit 12 uses the camera to measure the object 1 and its periphery 2 so that the four vertices of the periphery 2 of the object to be measured 1 and a plurality of lattice points (B11 to Bnm) on the object to be measured 1 are captured. To obtain a captured image. The pre-correction image coordinate acquisition unit 14 obtains image coordinates of the four vertices before tilt correction from the obtained captured image. The corrected image coordinate calculation unit 15 obtains the image coordinates of the four vertices after tilt correction based on the length between the four vertices. The projective transformation coefficient calculation unit 16 obtains a projective transformation coefficient for tilt correction based on the relational expression between the image coordinates of the four vertices before tilt correction and the image coordinates of the four vertices after tilt correction. The pre-correction grid point coordinate calculation unit 17 obtains image coordinates of a plurality of grid points (B11 to Bnm) before correction from the captured image. The corrected lattice point coordinate calculation unit 18 performs projective conversion on the image coordinates of the plurality of lattice points using the obtained projective conversion coefficient, thereby correcting the image of the plurality of lattice points (B11 to Bnm) after tilt correction. Find the coordinates.

  In this precision measuring device 10, by correcting the projection transformation coefficients for tilt correction for the four vertices designated in a square shape around the periphery 2 of the measurement object 1, the tilt correction without the need for correction by the user is realized. it can.

<Precise measurement method>
FIG. 7 is a flowchart of the precision measurement method according to the first embodiment. This precision measurement method will be described below.
(A) Four vertices are designated in a square shape in the periphery 2 of the object 1 to be measured, and at least 5 sides (for example, the circumference 3 of the rectangle) of the length between the vertices and the 4 sides and 2 diagonal sides of the rectangle Two sides and two diagonal sides or four sides around a quadrangle and one diagonal side) are measured (S01). FIG. 2 is a plan view showing four vertices 3 designated in a square shape around the object to be measured 1 and a plurality of lattice points 4 designated on the object to be measured 1. FIG. 3 is a modification of FIG. 2 and is a plan view in the case where the four vertices 3 form a quadrilateral other than a rectangle. FIG. 4 is a plan view in which a quadrangle composed of four vertices 3 around the periphery 2 of the measurement object 1 is extracted. It should be noted that the length can be measured more accurately by using a marker on the periphery 2 of the measurement object 1 as the vertex 3. Therefore, FIGS. 5A to 5J show examples of markers used as vertices. Moreover, it is preferable to designate the vertex 3 so as to form a quadrangular shape in the same plane. Note that the length between the vertices can be measured by a commonly used method, for example, measurement by a measure, laser measurement, or stereo imaging by two cameras.
Here, the measurement object 1 is, for example, concrete, mortar, or the like. In addition, the to-be-measured object 1 is not restricted to the said concrete, mortar, etc. Further, it is desirable that the periphery 2 of the measurement object 1 is a portion made of a material that is disposed in the vicinity of the measurement object 1 and is less likely to be displaced with time than the measurement object 1. The periphery 2 of the measurement object 1 may be, for example, a metal frame.

(B) Photographing the measurement object 1 and its surroundings 2 with a camera so that the four vertices 3 around the measurement object 1 and a plurality of designated lattice points 4 on the measurement object 1 can be seen. Thus, a photographed image is obtained (S02). FIG. 6 is a diagram illustrating an example of an image obtained by photographing four vertices. In this case, shooting can be performed with one camera. In the captured image, the points corresponding to the four vertices are α, β, γ, and δ, respectively. In this case, the camera lens distortion (barrel distortion, pincushion distortion, etc.) may be removed to obtain a captured image.
(C) Image coordinates of four vertices before tilt correction are obtained from the obtained captured image (S03). The image coordinates of the four vertices can be obtained in pixel units as described later.
(D) Based on the measured length between vertices, the image coordinates of the four vertices after tilt correction are obtained (S04). In this case, the image coordinates of the four vertices are based on the measured length (m), and are, for example, m units.
(E) Relationship between the image coordinates of the four vertices before the tilt correction obtained in the step (c) and the image coordinates of the four vertices after the tilt correction obtained in the step (d) Based on the equation, a coefficient for projective transformation for tilt correction is obtained (S05).
(F) Image coordinates of a plurality of grid points before tilt correction are obtained from the captured image (S06).
(G) The image coordinates of the plurality of grid points after tilt correction are obtained by projective transformation of the image coordinates of the plurality of grid points using the obtained projective transformation coefficient (S07).
As described above, the image coordinates of a plurality of grid points on the object to be measured 1 after tilt correction can be obtained.

Below, the detail of each step of this precision measuring method is demonstrated.
<About step (c)>
Points α, β, γ, and δ on the captured image corresponding to the four vertices (A, B, C, and D) in the real space are image coordinates obtained from the image, and Street.
α = (x ₁ , y ₁ )
β = (x ₂ , y ₂ )
γ = (x ₃ , y ₃ )
δ = (x ₄ , y ₄ )
The four points α, β, γ, and δ are image coordinates obtained by photographing four vertices, and are image coordinates before tilt correction.

以上によって、点Ｃの座標（ＡＥ、ＣＥ）が得られる。 <About step (d)>
・ In the case of 3 sides and 2 diagonal sides of a rectangle When measuring the length of 3 sides and 2 diagonal sides as the length between 4 vertices, the coordinates of each point of the 4 vertices are as follows Can be obtained in this way.
Among the quadrilateral ABCD formed by four points (A, B, C, D), each of the triangle ABC composed of two peripheral sides (AB, BC) and one diagonal side (AC) Among the points (A, B, C), the coordinates (AE, CE) of the point C are obtained. At this time, the direction from the vertex A to the vertex B is the X axis, and the vertex A is the origin.
First, regarding the triangle ABC, the sum of the three sides is 2 s, and the area S of the triangle ABC is obtained by Heron's formula as follows.

As described above, the coordinates (AE, CE) of the point C are obtained.

Next, the coordinates (AF, DF) of the point D among the points (A, B, D) of the triangle ABD composed of two peripheral sides (AB, DA) and one diagonal side (BD) are set. Ask. As described above, the coordinate system uses the direction from vertex A to vertex B as the X axis and vertex A as the origin.
For the triangle ABD, the sum of the three sides is 2 s, and the area S of the triangle ABD is obtained from the Heron formula as follows:

Therefore, the coordinates of each point of ABCD are as follows when the coordinate of point A is the origin.
A = (0, 0) ⇒ (X ₁ , Y ₁ )
B = (AB, 0) ⇒ (X ₂ , Y ₂ )
C = (AE, CE) => (X ₃ , Y ₃ )
D = (AF, DF) => (X ₄ , Y ₄ )

In the case of four sides and one diagonal side of the rectangle On the other hand, as the length between the four vertices, among the four sides and two diagonal sides of the rectangle, the case of the four sides and one diagonal side of the rectangle The case where it measured is demonstrated. For example, when measuring the four surrounding sides and the diagonal AC, the origin can be set to the point A and the diagonal AC can be set to the X-axis to perform the same handling as described above. For example, with respect to the triangle ABC, a perpendicular line is drawn from the point B to the diagonal AC, and a leg of the perpendicular line is taken as a point E. Thereafter, BE and AE are calculated using the Heron formula in the same manner as described above, and the image coordinates of the point B are obtained. For the triangle ACD, a perpendicular line is drawn from the point D to the diagonal AC, and the perpendicular leg is set to the point F. Thereafter, the image coordinates of the point D can be obtained by calculating DF and AF using the Heron formula in the same manner as described above.

なお、上述のように、あおり補正前の画像上の画像座標（ｘ_ｎ，ｙ_ｎ）はピクセル単位で得られる。一方、あおり補正後の画像座標（Ｘ_ｎ，Ｙ_ｎ）は、例えばｍ単位で得られる。それぞれの単位が相違するので、それぞれの単位系を対応させる任意の定数としてｋを設けた。ｋは、任意の定数であるが、決定の方法の一つとしては、例えば、以下の方法が挙げられる。
ｋ＝αβ間の距離（ｐｉｘｅｌ（画素））／ＡＢ間の距離（ｍ） <About step (e)>
The coordinate transformation (projective transformation) from the image coordinates (x _n , y _n ) on the image before tilt correction to the image coordinates (X _n , Y _n ) after tilt correction is shown as follows. In the following formula, a, b, c, d, e, f, A, and B are parameters in the projective transformation for tilt correction.

As described above, the image coordinates (x _n , y _n ) on the image before tilt correction are obtained in units of pixels. On the other hand, the image coordinates (X _n , Y _n ) after tilt correction are obtained in units of m, for example. Since each unit is different, k is set as an arbitrary constant to correspond to each unit system. k is an arbitrary constant, but one of the determination methods is, for example, the following method.
k = distance between αβ (pixel (pixel)) / distance between AB (m)

  Eight sets of equations can be obtained by substituting the image coordinates of the four vertices before and after the tilt correction into the above-described tilt correction projective transformation formula, and by solving this, parameters in the tilt correction projective transformation can be obtained. a, b, c, d, e, f, A, and B can be calculated.

<Step (f)>
Image coordinates of a plurality of grid points 4 on the measurement object 1 are obtained on the captured image.

<Step (g)>
Using the obtained parameters a, b, c, d, e, f, A, and B in the projective transformation for tilt correction, the projective transformation is performed on the image coordinates of a plurality of grid points 4 arranged on the object 1 to be measured. By doing so, it is possible to obtain image coordinates of a plurality of lattice points of the object 1 after tilt correction.

  In addition, the said step (a)-(g) is performed in multiple times, and the image of the some grid point 4 of the to-be-measured object 1 is obtained as an average of the image coordinate of the some grid point 4 of the to-be-measured object 1 calculated | required by each. Coordinates may be obtained.

<Computer program for precision measurement>
FIG. 8 is a flowchart of the precision measurement computer program according to the first embodiment.
(A) The measurement object 1 and the measurement object 1 so that the four vertices 3 designated in a square shape on the periphery 2 of the measurement object 1 and the plurality of lattice points 4 designated on the measurement object 1 are shown. Image coordinates of four vertices before tilt correction are obtained from a photographed image obtained by photographing the periphery 2 (S11).
(B) Image coordinates after tilt correction of the four vertices 3 based on the length between the four vertices 3 and the length measured for at least five of the four sides and two diagonal sides of the rectangle. Is obtained (S12).
(C) Relationship between the image coordinates of the four vertices 3 before the tilt correction obtained in step (a) and the image coordinates of the four vertices 3 after the tilt correction obtained in step (b) Based on the equation, a coefficient for projective transformation for tilt correction is obtained (S13).
(D) Image coordinates of a plurality of lattice points 4 on the object to be measured 1 are obtained on the photographed image (S14).
(E) Projective transformation is performed on the image coordinates of a plurality of grid points 4 of the measurement object 1 using the obtained projective transformation coefficients, so that a plurality of grid points 4 on the measurement object 1 after tilt correction is performed. The image coordinates are obtained (S15).
By causing the computer to execute the above steps, the image coordinates of the plurality of grid points 4 on the object to be measured 1 can be precisely measured.

  Further, the computer program for precision measurement may be stored in a computer-readable recording medium.

(Embodiment 2)
<Displacement measurement method>
FIG. 9 is a flowchart of a displacement measuring method for measuring the temporal displacement of the image coordinates of a plurality of grid points 4 on the object to be measured 1 from the first time to the second time according to the second embodiment. is there. Below, this displacement measuring method is demonstrated.
(1) At the first time, the following precision measurement method is performed to obtain the image coordinates of the plurality of lattice points 4 on the measurement object 1 at the first time (S21).
(1-a) Four vertices are designated in a square shape in the periphery 2 of the object 1 to be measured, and at least five sides (for example, a quadrangular shape) of the length between the vertices, the four sides of the quadrangle, and the two diagonal sides are specified. Three sides and two diagonal sides or four sides of a square and one side are measured (S31).
(1-b) The measurement object 1 and its surroundings are captured by the camera so that the four vertices 3 of the periphery 2 of the measurement object 1 and a plurality of designated grid points 4 on the measurement object 1 are shown. 2 is photographed to obtain a photographed image (S32).
(1-c) Image coordinates of four vertices before tilt correction are obtained from the obtained captured image (S33).
(1-d) Based on the measured length between vertices, the image coordinates of the four vertices after tilt correction are obtained (S34).
(1-e) Image coordinates of four vertices before tilt correction obtained in step (1-c) and images of four vertices after tilt correction obtained in step (1-d) Based on the relational expression between the coordinates, a coefficient for projective transformation for tilt correction is obtained (S35).
(1-f) Image coordinates of a plurality of grid points before tilt correction are obtained from the photographed image (S36).
(1-g) The image coordinates of the plurality of grid points after tilt correction are obtained by projective transformation of the image coordinates of the plurality of grid points using the obtained projective transformation coefficient (S37).

(2) In the second time after the first time, the following precision measurement method is performed to obtain the image coordinates of the plurality of lattice points 4 on the measurement object 1 in the second time (S22). .
(2-a) Four vertices are designated in a square shape around the periphery 2 of the measurement object 1, and at least five sides (for example, a quadrangular shape) of the length between the vertices, the four sides of the quadrangle, and the two diagonal sides are specified. Three surrounding sides and two diagonal sides or four square sides and one diagonal side are measured (S41).
(2-b) The measurement object 1 and its surroundings by the camera so that the four vertices 3 of the periphery 2 of the measurement object 1 and a plurality of designated grid points 4 on the measurement object 1 are shown. 2 is photographed to obtain a photographed image (S42).
(2-c) Image coordinates of four vertices before tilt correction are obtained from the obtained captured image (S43).
(2-d) Based on the measured length between vertices, the image coordinates of the four vertices after tilt correction are obtained (S44).
(2-e) Image coordinates of four vertices before tilt correction obtained in the step (2-c) and images of four vertices after tilt correction obtained in the step (2-d) Based on the relational expression between the coordinates, a projective transformation coefficient for tilt correction is obtained (S45).
(2-f) Image coordinates of a plurality of grid points before tilt correction are obtained from the photographed image (S46).
(2-g) Projective transformation is performed on the image coordinates of a plurality of grid points using the obtained projective transformation coefficient to obtain image coordinates of the plurality of grid points after tilt correction (S47).
(3) The time-dependent displacement from the image coordinates of the plurality of lattice points on the object to be measured at the first time to the image coordinates of the plurality of lattice points on the object to be measured at the second time is obtained. (S23).
As described above, it is possible to measure the temporal displacement of the image coordinates of the plurality of grid points on the object to be measured from the first time to the second time.
By specifying a plurality of lattice points on the object to be measured 1 by this displacement measuring method, for example, displacement due to the occurrence of a crack on concrete can be measured over a plane.

  In FIG. 9, a step (S21) of obtaining image coordinates of a plurality of grid points 4 on the measurement object 1 at the first time, and a plurality of grid points on the measurement object 1 at the second time. The step (S22) of obtaining the image coordinates of 4 and the step (S23) of obtaining the temporal displacement of the image coordinates of the plurality of lattice points on the object to be measured from the first time to the second time are described. It has only been done. Steps (1-a) (S31) to (1-g) (S37) and (2-a) (S41) to (2-g) (S47) are respectively performed in steps (a) ( Since it is substantially the same as S01) to (g) (S07), its description is omitted.

  In the above case, the camera may be fixed from the first time to the second time. Thereby, in the step of obtaining the image coordinates of the plurality of grid points 4 on the object to be measured 1 in the second time, steps (b) (S42), (f) (S46), (g ) (S47) only, the image coordinates of the plurality of grid points 4 on the object to be measured 1 can be obtained. In other words, since the camera is fixed from the first time to the second time, it can be considered that the coefficient of the projective transformation for tilt correction is substantially the same between the first time and the second time. The tilt correction steps (a) (S41) and (c) (S43) to (e) (S45) at the second precision measurement can be omitted.

<Computer program for displacement measurement>
FIG. 10 is a flowchart of the precision measurement computer program according to the first embodiment.
(1) In the first time, the following precision measurement method is performed to obtain image coordinates of a plurality of lattice points 4 on the measurement object 1 in the first time (S51).
(1-a) The object to be measured so that the four vertices 3 specified in a square shape on the periphery 2 of the object to be measured 1 and the plurality of lattice points 4 specified on the object to be measured 1 are shown. Image coordinates of four vertices before tilt correction are obtained from captured images obtained by photographing 1 and its periphery 2 (S61).
(1-b) Based on the length between four vertices 3 and the length measured for at least five of the four sides and two diagonal sides of the quadrangle, Image coordinates are obtained (S62).
(1-c) Image coordinates of the four vertices 3 before the tilt correction obtained in the step (1-a) and images of the four vertices 3 after the tilt correction obtained in the step (1-b) Based on the relational expression between the coordinates, a projective transformation coefficient for tilt correction is obtained (S63).
(1-d) Image coordinates of a plurality of grid points 4 on the object to be measured 1 are obtained on the photographed image (S64).
(1-e) Projective transformation is performed on the image coordinates of a plurality of grid points 4 of the measurement object 1 using the obtained projective transformation coefficient, thereby a plurality of grids on the measurement object 1 after tilt correction. Image coordinates of the point 4 are obtained (S65).

(2) In the second time after the first time, the following precision measurement method is performed to obtain the image coordinates of the plurality of lattice points 4 on the measurement object 1 in the second time (S52). .
(2-a) The object to be measured so that the four vertices 3 specified in a square shape on the periphery 2 of the object to be measured 1 and the plurality of grid points 4 specified on the object to be measured 1 can be seen. Image coordinates of four vertices before tilt correction are obtained from the photographed image obtained by photographing 1 and its periphery 2 (S71).
(2-b) Based on the length between the four vertices 3 and the length measured for at least five of the four sides and the two diagonal sides of the quadrangle, Image coordinates are obtained (S72).
(2-c) Image coordinates of the four vertices 3 before the tilt correction obtained in the step (2-a) and images of the four vertices 3 after the tilt correction obtained in the step (2-b) Based on the relational expression between the coordinates, a projective transformation coefficient for tilt correction is obtained (S73).
(2-d) Image coordinates of a plurality of grid points 4 on the object to be measured 1 are obtained on the photographed image (S74).
(2-e) Projective transformation is performed on the image coordinates of the plurality of grid points 4 of the measured object 1 by using the obtained projective transformation coefficient, and a plurality of grids on the measured object 1 after tilt correction. The image coordinates of the point 4 are obtained (S75).
(3) Over time from the image coordinates of the plurality of grid points 4 on the measurement object 1 at the first time to the image coordinates of the plurality of grid points 4 on the measurement object 1 at the second time The displacement is obtained (S53).
By causing the computer to execute the above steps, it is possible to measure the temporal displacement of the image coordinates of the plurality of grid points 4 on the measurement object 1 from the first time to the second time.

  In FIG. 10, a step (S51) of obtaining image coordinates of a plurality of grid points 4 on the measurement object 1 at the first time, and a plurality of grid points on the measurement object 1 at the second time. The step (S52) of obtaining the image coordinates of 4 and the step (S53) of obtaining the temporal displacement of the image coordinates of the plurality of lattice points on the object to be measured from the first time to the second time are described. It has only been done. Steps (1-a) (S61) to (1-e) (S65) and (2-a) (S71) to (2-e) (S75) are respectively performed in steps (a) ( Since it is substantially the same as S11) to (e) (S15), its description is omitted.

  In the above case, the camera may be fixed from the first time to the second time. Thereby, in the step of obtaining the image coordinates of the plurality of lattice points 4 on the object to be measured 1 in the second time, the steps (2-d) (S74) and (2-e) (S75) in the precision measurement method described above. ) Only, the image coordinates of the plurality of grid points 4 on the object to be measured 1 can be obtained. In other words, since the camera is fixed from the first time to the second time, it can be considered that the coefficient of the projective transformation for tilt correction is substantially the same between the first time and the second time. The tilt correction steps (a) (S71) to (c) (S73) at the second precision measurement can be omitted.

  Further, the displacement measuring computer program may be stored in a computer-readable recording medium.

  In the precision measurement device, precision measurement method, precision measurement computer program, displacement measurement method, and displacement measurement computer program according to the present invention, it is sufficient to designate four vertices around the object to be measured as a quadrangular shape. Therefore, it is possible to perform precise measurement with a high degree of freedom.

1 Object 2 Perimeter 3 Vertex (A1 to A4)
4 Lattice points (B11 to Bnm)
DESCRIPTION OF SYMBOLS 10 Precision measuring device 11 Measuring part 12 Image | photographing part 13 Control part 14 Image coordinate acquisition part before correction | amendment 15 Image coordinate calculation part 16 after correction | amendment Projection conversion coefficient calculation part 17 Lattice point coordinate calculation part 21 CPU
22 memory 23 storage means 24 input / output means 25 display means 26 communication means

Claims (9)

(A) For the four vertices designated in a quadrangular shape around the object to be measured, the length between the four vertices is the length of at least five of the four sides and two diagonal sides of the quadrilateral. Measuring unit to obtain
(B) Photographing the object to be measured and its surroundings with a camera so that the four vertices around the object to be measured and a plurality of lattice points designated on the object to be measured are photographed. An imaging unit for obtaining images;
(C) a pre-correction image coordinate acquisition unit for obtaining image coordinates of the four vertices before tilt correction from the obtained captured image;
(D) a corrected image coordinate calculation unit for obtaining image coordinates of the four vertices after tilt correction based on the measured length between the vertices;
(E) Projection transformation coefficient calculation for obtaining a projection transformation coefficient for tilt correction based on a relational expression between the image coordinates of the four vertices before tilt correction and the image coordinates of the four vertices after tilt correction And
(F) a pre-correction lattice point coordinate calculation unit for obtaining image coordinates of the plurality of lattice points in the captured image;
(G) For a plurality of grid points on the object to be measured, the image coordinates of the plurality of grid points are subjected to projective transformation using the obtained projective transformation coefficient, so that the measurement after tilt correction is performed. A grid point coordinate calculation unit for obtaining image coordinates of the plurality of grid points on an object;
Including precision measuring equipment. (A) A step of designating four vertices around the object to be measured in a quadrilateral shape, and measuring at least five sides out of four sides and two diagonal sides of the quadrangle as a length between the four vertices. When,
(B) Photographing the object to be measured and its surroundings with a camera so that the four vertices around the object to be measured and a plurality of lattice points arranged on the object to be measured are photographed. Obtaining an image;
(C) obtaining image coordinates of the four vertices before tilt correction from the obtained captured image;
(D) obtaining image coordinates of the four vertices after tilt correction based on the measured length between the vertices;
(E) Between the image coordinates of the four vertices before the tilt correction obtained in the step (c) and the image coordinates of the four vertices after the tilt correction obtained in the step (d) Obtaining a projective transformation coefficient for tilt correction based on the relational expression of
(F) obtaining image coordinates of the plurality of grid points in the captured image;
(G) For a plurality of grid points arranged on the object to be measured, the image coordinates of the plurality of grid points are subjected to projective transformation using the obtained projective transformation coefficient, and after the tilt correction, Obtaining image coordinates of the plurality of grid points on the object to be measured;
Including precision measurement methods.   The step (a) measures at least one of four sides and one diagonal side of the quadrangle or three sides and two diagonal sides of the quadrangle as the length between the vertices. The precision measurement method described in 1. (A) Photographing the object to be measured and its surroundings so that four vertices designated in a square shape around the object to be measured and a plurality of lattice points arranged on the object to be measured can be seen. Obtaining the image coordinates of the four vertices before tilt correction from the captured image obtained by
(B) Based on the measured length between the four vertices and the length measured on at least five sides of the four sides and the two diagonal sides of the quadrangle, the four after the tilt correction Obtaining image coordinates of vertices;
(C) Between the image coordinates of the four vertices before the tilt correction obtained in the step (a) and the image coordinates of the four vertices after the tilt correction obtained in the step (b) Obtaining a projective transformation coefficient for tilt correction based on the relational expression of
(D) Using the obtained projective transformation coefficient, the coordinates of the plurality of grid points arranged on the object to be measured are projectively transformed so that the tilt-corrected object on the object to be measured is corrected. Obtaining coordinates of the plurality of grid points arranged;
A computer program for precision measurement that causes a computer to execute each step of and precisely measures the coordinates of a plurality of lattice points of an object to be measured.   A computer-readable recording medium storing the precision measurement computer program according to claim 4. Performing the precision measurement method according to claim 2 at a first time to obtain image coordinates of the plurality of lattice points on the object to be measured at the first time;
In the second time after the first time, the precision measurement method according to claim 2 is performed, and image coordinates of the plurality of lattice points on the measurement object in the second time are obtained. Seeking steps,
Displacement over time from the image coordinates of the plurality of lattice points on the object to be measured at the first time to the image coordinates of the plurality of lattice points on the object to be measured at the second time A step of seeking
Displacement measuring method. The camera is fixed from the first time to the second time,
In the step of obtaining image coordinates of the plurality of lattice points on the object to be measured at the second time, only steps (b), (f), and (g) in the precision measurement method according to claim 2. The displacement measurement method according to claim 6, wherein image coordinates of the plurality of lattice points on the object to be measured in the second time are obtained. A computer executes each step of the displacement measuring method according to claim 6 or 7,
A computer program for displacement measurement that measures a time-dependent displacement of image coordinates of a plurality of lattice points on an object to be measured over a first time to a second time.   A computer-readable recording medium storing the displacement measurement computer program according to claim 8.

Images