JPH0791928A - Method for measuring three-dimensional shape - Google Patents

Abstract

PURPOSE:To easily adjust the position of a measuring head and perform highly precise measurement. CONSTITUTION:An object to be measured is moved along a carrying line A and three-dimensional shape of the object to be measured is measured by a light cutting method by means of a measuring head 4 provided with a slit light source 7 and a teleccamera 8. In a plurality of postures of an exclusive correction jig 17, the positions in a measuring head coordinate system of a specified part of the jig 17 are measured with the measuring head 4. The measurement result is data-processed, so that a dislocation of an angle of the coordinate system of the carrying line A and the measuring head 4, and a correction formula for correcting the dislocation are found and measured data of the object to be measured are corrected by means of the correction formula.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION This invention uses a light-section method.
The present invention relates to a method for measuring a three-dimensional shape, for example, a method for measuring a three-dimensional shape of a required portion of a vehicle body which is moved along a transportation line in an automobile manufacturing line.

[0002]

2. Description of the Related Art As a measuring method of this kind, a measuring head having a slit light source and a two-dimensional image pickup device such as a television camera is provided at a predetermined position around a conveying line, and an object to be measured is conveyed along the conveying line by a conveying means. The slit light source irradiates the measured object with slit light at a plurality of measured object moving positions, and the two-dimensional imaging device captures the optical cutting line formed by the slit light hitting the surface of the measured object. The two-dimensional image data of the optical cutting line imaged by the two-dimensional image pickup device is converted into three-dimensional position data in the three-dimensional measuring head coordinate system of the measuring head by using the conversion position data, and the three-dimensional image data at a plurality of measured object moving positions is obtained. It is known to combine three-dimensional position data to obtain a three-dimensional shape of a required portion of the measured object.

[0003]

In the above measuring method, the conversion position data is the three-dimensional measuring head of the measuring head including the two-dimensional image data of the light cutting line in the screen coordinate system of the two-dimensional image pickup device including the slit light surface. It is for converting into three-dimensional position data in the coordinate system, and is obtained by calibration prior to measurement. Further, the above-mentioned measurement is premised on that the transport line of the object to be measured and one coordinate axis of the measuring head coordinate system of the measuring head are completely parallel. The coordinate value related to one coordinate axis of the measuring head coordinate system is given by the amount of movement of the measured object on the transport line. For this reason, if the transport line and one coordinate axis of the measuring head coordinate system are not perfectly parallel, an error will occur in the measured value, and these must be perfectly matched in order to perform highly accurate measurement. is there. Therefore, strict position adjustment is required when the measurement head is installed at the measurement site, but this adjustment is very difficult and time-consuming.

The object of the present invention is to solve the above problems,
An object of the present invention is to provide a three-dimensional shape measuring method in which the measurement head can be easily adjusted and highly accurate measurement can be performed.

[0005]

According to the three-dimensional shape measuring method of the present invention, a measuring head provided with a slit light source and a two-dimensional image pickup device is provided at a predetermined position around a conveying line, and an object to be measured is conveyed by the conveying means. Two-dimensional imaging of the optical cutting line formed by moving along the transport line, irradiating the slit light from the slit light source to the DUT at a plurality of moving positions of the DUT, and hitting the slit light on the surface of the DUT. A plurality of objects to be measured are obtained by converting the two-dimensional image data of the optical cutting line imaged by the device and the three-dimensional position data in the coordinate system of the three-dimensional measuring head of the measuring head using the conversion position data. In the three-dimensional shape measuring method for synthesizing the three-dimensional position data at the moving position to obtain the three-dimensional shape of the required portion of the measured object, a dedicated correction jig is used and The position of the predetermined part of the jig in the measuring head coordinate system is measured using a measuring head in a plurality of postures of the jig, and the measurement result is data-processed to determine the angle between the transport line and one coordinate axis of the measuring head coordinate system. Deviation and a correction equation for correcting this deviation, and using this correction equation, 3 at each moving position of the object to be measured.
The feature is that the dimensional position data is corrected.

For example, the correction jig is provided with a base fixed to the transfer means and moved along the transfer line, and a rod-shaped member to be measured which extends substantially parallel to the transfer line in the reference posture. , For the base,
The moving unit is moved in the length direction of the reference posture, and the rotating unit is rotated about two axes orthogonal to the moving direction.
The base of this correction jig is fixed to the transfer means, and the transfer means moves the member to be measured with respect to the base while the transfer means is stopped. The measuring member is irradiated with slit light, and the light cutting line formed by the slit light hitting the surface of the member to be measured is imaged by the two-dimensional imaging device, and the two-dimensional image data of the light cutting line imaged by the two-dimensional imaging device is obtained. Based on this, the posture of the member to be measured in the reference posture with respect to one coordinate axis of the measuring head coordinate system is obtained, and the correction jig is moved to the conveying line by the conveying means while the member to be measured is fixed in the reference posture with respect to the base. The slit light source irradiates the member to be measured with slit light at a plurality of jig moving positions, and the slit light strikes the surface of the member to be measured. Capturing a line in two-dimensional imaging device,
Based on the two-dimensional image data of the optical cutting line imaged by the two-dimensional image pickup device and the posture of the member to be measured in the reference posture with respect to one coordinate axis of the measurement head coordinate system, which is obtained in advance, 1 of the conveyance line and the measurement head coordinate system Find the angle deviation from one coordinate axis and the correction formula.

[0007]

With the use of the dedicated correction jig, the position of the predetermined portion of the jig in the measuring head coordinate system is measured using the measuring head in a plurality of predetermined postures of the jig, and the measurement result is processed as data. As a result, a deviation of the angle between the transport line and one coordinate axis of the measuring head coordinate system and a correction formula for correcting this deviation are obtained, and the three-dimensional position data at each moving position of the object to be measured is calculated using this correction formula. Since the correction is performed, even if there is an angle deviation between the transport line and one coordinate axis of the measuring head coordinate system, a measurement error does not occur and the three-dimensional shape of the measured object can be accurately measured. .

[0008]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a conveyor (conveying means) (1) in an automobile manufacturing line, and an object to be measured such as a car body of an automobile which is moved by the conveyor (1) along a substantially horizontal constant conveying line (A). The structure of the main part of the three-dimensional shape measuring device (3) for measuring the three-dimensional shape of the required part of the object (2) is shown. Further, FIG. 2 shows a configuration of a main part of the measuring device (3). In the following description, front, rear, left and right refer to the moving direction of the DUT (2), that is, the direction of the transfer line (A).

The measuring device (3) comprises a first measuring head (4), a second measuring head (5) and a data processing device (6). The two heads (4) and (5) are arranged on the left and right sides of the transfer line (A) so as to face each other. First head
(4) has a slit light source (7) and a television camera (8). The slit light source (7) uses an LD (laser diode), a cylindrical lens, etc., not shown,
It is for irradiating the DUT (2) with slit light (9). In this case, the slit light surface (9a) is substantially vertical and substantially orthogonal to the carrier line (A), and the first three-dimensional measuring head is composed of three coordinate axes of X _h1 axis, Y _h1 axis and Z _h1 axis which are orthogonal to each other. The coordinate system is set. Z _h1 axis is slit light surface
Vertical coordinate axis parallel to (9a), X _h1 axis is slit light surface
It is a horizontal coordinate axis parallel to (9a). The Y _h1 axis is a coordinate axis in the front-back direction orthogonal to the slit light surface (9a), and is parallel or almost parallel to the transport line (A). The television camera (8) images the light cutting line (10) formed by the slit light (9) hitting the surface of the DUT (2) and outputs the two-dimensional image data to the data processing device (6). One, two
A three-dimensional imaging device is configured. The second head (5) is the first
Like the head (4), it has a slit light source (11) and a television camera (12), which are arranged symmetrically to those of the first head (4). In FIG. 1, (13) is a light source
Slit light from (11), (13a) is the slit light plane, (1
Reference numeral 4) is a light cutting line formed by the slit light 13 coming into contact with the surface of the DUT 2. For the second head (5), as in the first head coordinate system, a second three-dimensional head measurement coordinate system including three coordinate axes of the X _h2 axis, the Y _h2 axis and the Z _h2 axis which are orthogonal to each other is set. There is. These two head coordinate systems are also symmetrical. Then, as will be described later in detail, the first slit light surface (9a) and the second slit light surface (1a
3a) matches, X _h1 axis and X _h2 axis coincides, Y _h1 axis and Y _h2
The positions and postures of the two heads (4) and (5) are adjusted so that the axes thereof are parallel to each other and the Z _h1 axis and the Z _h2 axis are parallel to each other. Note that, in FIG. 1, due to the space, the X _h1 axis, the Y _h1 axis, the Z _h1 axis, the X _h2 axis, the Y _h2 axis, and the Z _h1 axis.
_{The h2} axis is shown translated as appropriate. Further, in the state where the two heads (4) and (5) are adjusted as described above, the X _h1 axis (X _h2 axis), the Y _h1 axis (Y _h2 axis) and the Z _h1 axis (Z _h2 Axes) are called X _h axis, Y _h axis and Z _h axis. If there is no confusion with other coordinate systems, X
_{The h} axis, the Y _h axis, and the Z _h axis may be simply referred to as the X axis, the Y axis, and the Z axis.

The data processing device (6) is the same as the device under test (2).
Data processing is performed for measuring the dimensional shape, and is configured by a microcomputer or the like. Figure 2
As shown in, the light sources (7) and (11) of the two heads (4) and (5) are also connected to the data processing device (6). Although detailed illustration is omitted, the conveyor (1) is driven to drive the DUT.
A conveyor drive device (15) equipped with an electric motor for moving (2), and, for example, by detecting the drive amount of the drive device (15), the object to be measured (2) on the conveyor (1)
A line movement amount detector (16) that detects the movement amount (line movement amount) etc. is provided, and the drive device (15) and line movement amount detector (16) are also connected to the data processing device (6). Has been done.

Prior to the measurement, each head (4) and (5) was calibrated by a television camera by calibration.
(8) The converted position data for converting the two-dimensional image data of the light cutting line in the screen coordinate system into the three-dimensional position data in the head coordinate system is obtained, and the data processing device is obtained.
It is stored in the memory of (6). Also two heads
(4) The position adjustment and the head coordinate system correction of (5) are performed. Then, as will be described in detail later, using a dedicated correction jig (17) (see FIG. 3), the deviation of the angle of the transfer line (A) with respect to the Y _h axis of the measurement coordinate system and the deviation thereof are corrected. A correction formula for this is calculated, and this is stored in the memory of the data processing device (6).

At the time of measurement, the object to be measured (2) is continuously or intermittently moved at a predetermined speed along the transfer line (A) by the conveyor (1). Then, at a plurality of measured object moving positions, the measured object (2) from the light source (7) (11)
The slit light (9) (13) is irradiated to the slit light (9) (13), and the slit light (9) (13) hits the surface of the DUT (2) to form a light cutting line (10).
Take images of (14) with the TV cameras (8) and (12).
(8) Convert the two-dimensional image data of the optical cutting line imaged in (12) into three-dimensional position data in the head coordinate system by using the conversion position data, and synthesize the three-dimensional position data in the plurality of measured object moving positions. Then, the three-dimensional shape of the required portion of the DUT (2) is obtained. At this time, the amount of movement of the DUT (2) detected by the line movement amount detector (16) and the data processing device
Using the correction formula stored in (6), the three-dimensional position data at each transport position is corrected, and the three-dimensional shape of the object to be measured (2) is obtained based on the corrected three-dimensional position data.

The adjustment of the positions of the two heads (4) and (5) and the correction of the head coordinate system are performed as follows (see the flowchart of FIG. 12).

In FIG. 12, first, two heads (4)
(5) is installed horizontally (step 101), and the Z-axis directions of the two head coordinate systems are aligned. Then two heads (4)
Align the optical axes of the slit light (9) and (13) of (5) (step 102
) Match the ZX planes of the two head coordinate systems. Next, the positions of the slit end points at the same height (Z coordinate value) position in the head coordinate system are measured, and these are designated as (x ₁ , x ₂ ) and (z ₁ , z ₂ ) respectively (step 103).
). In addition, the slit endpoint refers to any one of predetermined endpoints of the two-dimensional image of the light cutting line of the slit light imaged by the television camera. Then, finally, the difference between the heights of both slit end points is calculated by the following equation to obtain a correction value dz in the Z-axis direction (step 104), and the XY planes of the two head coordinate systems are made to coincide with each other.

Z ₁ -z ₂ → dz In the above correction, the two heads (4) and (5) are preliminarily arranged in each head (4).
(5) Calibrate each unit, and measure each slit light surface.
The two head coordinate systems are made to coincide with each other in terms of hardware and software on the assumption that they have their own coordinate system (head coordinate system) for (9a) and (13a).

The reason why the three-dimensional position data is corrected using the correction formula is as follows. The head coordinate system of the two heads (4) and (5) is adjusted by the correction between the two heads (4) and (5) described earlier in the flowchart of FIG. 12, and the second head ( Since 5) is for obtaining the measurement start point in this coordinated head coordinate system, it is not necessary to obtain the correction formula for the second head (5), but only for the first head (4). You should ask. Therefore, the following description relates to the first head (4).

3 converted by using the conversion position data from the two-dimensional image data of the TV camera (8) of the measuring head (4)
Dimensional position data is in the head coordinate system and is 3
The coordinate value about the Y _h axis used when synthesizing the three-dimensional position data to obtain the three-dimensional shape is given by the movement amount (line movement amount) of the object to be measured (2) on the transfer line (A).
Therefore, if the transport line (A) and the Y _h axis are completely coincident with each other or are completely parallel to each other, the 3D position data converted from the 2D image data is combined as it is to obtain 3
The measurement error does not occur even if the dimensional shape is obtained. On the other hand, when the angle between the transport line (A) and the Y _h axis is misaligned, the 3D position data converted from the 2D image data is directly combined to obtain the 3D shape. Occurs. Therefore, to prevent measurement error, 2
The three-dimensional position data converted from the three-dimensional image data is corrected using the correction formula.

Next, referring to FIG. 5, the three-dimensional position data converted from the two-dimensional image data is combined as it is to generate 3
The reason why a measurement error occurs due to the deviation of the angle between the transport line (A) and the Y _h axis when the three-dimensional shape is obtained will be described.

Here, first, the actual posture angles are represented by θ _x and θ.
when is _z, consider whether the head coordinate system (XYZ) _h and the conveying line coordinate system (XYZ) deviation of _l is the measured value θ _x ', θ _z' affects how respect. Line coordinate system, X _l axis, and is composed of three coordinate axes orthogonal to each other Y _l-axis and Z _l axis. Z _l axis in the vertical direction coordinate axis, X _l axis in the lateral direction coordinate axis, Y _l axis conveying line
It is a coordinate axis in the front-back direction that matches (A).

The line coordinate system is X _{h with} respect to the head coordinate system.
When considering a state in which the posture is φ around the axis and ψ around the Z _h axis, the unit vector e _l (Y _h
The direction vector of the axis) is expressed by the following equation (1).

[0022]

[Equation 1] However, Rot (φ) _x , Rot (ψ) _z and _ey are as follows.

[0023]

[Equation 2] Further, the direction vector of the ridge to be measured is e, the angle around the Z _l axis with respect to e _l is θ _z , and the angle around the X _l axis is θ _x.
And Further, the change amount (dx, dz) of the slit end point observed at this time is set as the displacement vector Δ.

When θ _x = θ _z = 0 ° (reference posture), θ
_x 'and θ _z ' are as follows. In this case, since the posture angle of the ridgeline in the state where e and e _l are coincident with each other is measured, it is observed that the displacement vector Δ = 0 and θ _x ′ = θ _z ′ = 0 °.

When θ _x and θ _z ≠ 0 °, θ _x 'and θ _z '
Is as follows. For simplicity, let us consider the case where θ _x = 0 °, that is, the ridge line is rotated θ _z = θ only around the Z _l axis. At this time, the direction vector e of the ridge line is described in the head coordinate system (XYZ) _{h as} in the following equation (2).

[0026]

[Equation 3] However, Rot (θ _z ) is as follows.

[0027]

[Equation 4] Now, the displacement of the slit end point when the line movement amount is L is Δ
Then, in FIG. 5, O ′ P = L.

Therefore, the vector O'P becomes:

[0029]

[Equation 5] Then, from the equation (1), OX, O ′ O and OZ are as follows.

OX = -L.sin .psi..cos φ O'O = L.cos .psi..cos φ OZ = L.sin φ Further, according to the ratio (correlation) between the vector O'Q and e,
Using (2), the following relation holds.

[0031]

[Equation 6] Therefore, the following equation is obtained.

[0032]

[Equation 7] Here, θ _z 'and θ _x ' are defined by the following equations.

Tan θ _z ′ = −dx / L tan θ _x ′ = dz · cos θ _z ′ / L Therefore, tan θ _z ′ and tan θ _x ′ are as follows.

[0034]

[Equation 8] When θ _z ′ and θ _x ′ are calculated using the above formula, the results shown in Table 1 below are obtained.

[0035]

[Table 1] Here, for simplicity, the case where the line coordinate system has no rotation around the Y _h axis is considered. When considering the rotation around the Y _h axis, the result varies depending on the order of the rotation axes, but first the rotation is ω around the Y _h axis, then φ around the X _h axis, and finally ψ around the Z _h axis. When considering the posture state, e _l and e
Are respectively expressed by the following equations, and the same consideration can be made.

[0036]

[Equation 9] However, Rot (ω) _y is as follows.

[0037]

[Equation 10] As described above, if there is a deviation between the head coordinate system of the measuring head and the line coordinate system of the transport line, it is understood that the measured value has non-linearity instead of the relative angular deviation. Therefore, in order to confirm the validity of the measurement angle in a different coordinate system, it is necessary to detect the posture of the coordinate system and make a correction in consideration of this.

In the above measuring device (3), a correction jig (17) as shown in FIG. 3 is used in order to obtain a correction formula for performing such correction.

The jig (17) includes a base (18) fixed horizontally on the conveyor (1), a movable base (19) mounted on the base (18) and movable in the horizontal direction, and a movable base. A first swivel base (20) attached to the (19) swivel centering on a vertical axis orthogonal to the moving direction of the mobile swivel (19), and a first swivel base attached to the first swivel base (20) ( 20) A second swivel base (21) swiveled about a horizontal axis orthogonal to the swivel center axis, and a rod-shaped beam (measuring member) whose base end is fixed to the second swivel base (21) ( 22). Transfer base (1) to base (18)
The axis in the moving direction of 9) is the Y axis, and the vertical axis orthogonal to the Y axis is Z.
The horizontal axis orthogonal to the axis, the Y axis and the Z axis is defined as the X axis. A reference coordinate system (jig coordinate system) (XYZ) is constituted by these X-axis, Y-axis and Z-axis. The central axis of rotation of the first swivel base (20) is always parallel to the Z axis. Also, the second swivel base
The turning center axis of (21) is always orthogonal to the Z axis. In a fixed reference posture, the central axis of rotation of the second swivel base (21) is parallel to the X axis, and the beam (22), especially its lower edge, is parallel to the Y axis.
In the base (18), the beam (22) in the standard posture is attached to the transfer line (Y
It is fixed on the conveyor (1) so that it is substantially parallel to the ( _l- axis) and the Z-axis is substantially vertical. Then, in such a state, the work for obtaining the correction formula is performed. Although detailed illustration is omitted, the jig (17) includes a movable table driving device (23) for moving the movable table (19), a movable table (19), that is, a beam.
Beam movement amount detector (24) for detecting the amount of movement of (22) in the Y-axis direction, first swivel drive device (25) for swiveling the first swivel base (20), first swivel base A first turning angle detector (rotary encoder) for detecting the turning angle of (20) (26),
Second swivel drive device for swiveling the second swivel base (21)
(27) and a second swivel angle detector (28) for detecting the swivel angle of the second swivel base (21) are provided, and these are connected to the data processing device (6) (Fig. 2).

When performing the work for obtaining the correction formula, slit light is emitted from the light source (7) of the first head (4) to the beam (22) of the jig (17).
The (9) is irradiated, and the light cutting line (29) formed by hitting the surface of the beam (22) is imaged by the television camera (8). FIG. 4 shows an example of an image of the light section line (29) (light section line image) at this time. In FIG. 4, the light section line image (3
The lower end (30a) of 0) is the slit end point. Although only the light section line image (30) is actually obtained, the upper and lower edges of the beam (22) are shown by chain lines in FIG.

Next, with reference to the flow chart of FIG. 13, the operation for obtaining the above correction formula will be described.

In FIG. 13, first, the attitude matrix R of the reference coordinate system with respect to the head coordinate system is obtained (step 1). Next, the traveling direction vector e _l of the transport line (A) with respect to the head coordinate system is obtained (step 2). And
A correction formula is obtained based on this, and the coefficient (correction coefficient) is stored in the memory of the data processing device (6) as correction data (step 3).

Next, with reference to FIGS. 6 to 9, a method for obtaining the posture matrix R of the reference coordinate system in step 1 of FIG. 13 will be described.

In FIG. 6, the displacement vector of the slit end point when the beam (22) is rotated + θ about the Z axis from the reference posture (slit end point P ₀ ) is Δ ₁ , and the displacement vector when it is also rotated by −θ. Let Δ be Δ ₂
It is expressed as (3) and (4).

[0045]

[Equation 11] These are obtained as measured values.

On the other hand, the distances from the rotation center O'of the beam (22) to the slit end points P ₀ , P ₁ , P ₂ are L ₀ , L
_{If 1} and L ₂ , Δ ₁ and Δ ₂ are expressed by the following equations (5) and (6).

[0047]

[Equation 12] However, Rot (θ _z ), Rot (−θ _z ), R, E,
e _y is as follows.

[0048]

[Equation 13] Here, L ₀ is given as the amount of movement of the beam (22).

Therefore, equations (3), (5) and (4),
From (6), we obtain the following relation.

[0050]

[Equation 14] Therefore, we obtain the following equation.

[0051]

[Equation 15] However, x ₁ , x ₂ , y ₁ , and y ₂ are as follows.

X ₁ = -L ₁ · sin θ x ₂ = -L ₂ · sin θ y ₁ = L ₁ · cos θ-L ₀ y ₂ = L ₂ · cos θ-L ₀ Further, this equation is solved to obtain the following equation. (7), (8), (9)
To get

[0053]

[Equation 16] Here, R = Rot (ψ) _z · Rot (φ) _x · Rot
If (ω) _y is set, each element of R is as follows.

L _x = cos ω · cos ψ−sin φ ·
_{sinω · sinψ m x = cosω ·} sinψ + sinφ · sinω · c
_{osψ n x = -cosφ · sinω l} y = -cosφ · sinψ m y = cosφ · cosψ n y = sinφ l z = sinω · cosψ + sinφ · cosω · s
in ψ m _z = sin ω · sin ψ−sin φ · cos ω · c
os ψ _nz = cos φ · cos ω By substituting the equations (7), (8), and (9) into these, the posture angles φ, ω, and ψ around each axis are calculated, and each element of R is further obtained.

Further, the following relationship is obtained from FIG.

L ₁ ² + _Lo ² − | Δ ₁ | ² = 2 · L ₁
・ L ₀ · cos θ L ₂ ² + _Lo ² − | Δ ₂ | ² = 2 · L ₂ · L ₀ · co
sθ Then, L ₁ and L ₂ used in equations (5) and (6) are calculated by the following equations.

[0057]

[Equation 17] From the above, it can be seen that the attitude matrix R of the reference coordinate system with respect to the head coordinate system is determined from the measured values of Δ ₁ and Δ ₂ .

If the head coordinate system satisfies the above-mentioned conditions, the beam (22) is perpendicular to the Z axis in the standard posture, and therefore the beam (22) rotating around the Z axis forms one plane. That is, the locus of the intersection (slit end point) of the beam (22) and the slit light surface (9a) is a straight line. However, it is observed that the actually measured loci of the slit end points deviate from the straight line which should be actually due to the error due to quantization and mechanical factors (see FIG. 8).

Therefore, if the measured values (Δ _x1 , 0, Δ _z1 ) and (Δ _x2 , 0, Δ _z2 ) are used as the values of the displacement vectors Δ ₁ and Δ ₂ , the detection accuracy of R is inferior. Depending on the situation, a calculation error may occur.

Therefore, as a countermeasure, the following operation is added under the above assumption.

-(N-1) around the Z axis with respect to the reference posture
・ The posture is rotated by a constant angle θ from the θ / 2 posture (initial posture), and the displacement vector Δ _i ′ (i =
1 to n) are measured.

The locus of the measured displacement vector Δ _i 'is approximated to the following straight line by the method of least squares.

Z = a _z · x + b _z The x component dx _i of each displacement vector is substituted into the above equation to correct dz _i as in the following equation to reduce the error in the z component due to quantization.

A _z · dx _i + b _z → dz _{i Using the} corrected displacement vector Δ _i ′, L
Calculate _i .

[0065]

[Equation 18] However, θ _i is as follows.

[0066]

[Formula 19] Further, here, x _i ′ = L _i · cos θ _i , y _i ′ =
Letting L _i · sin θ _i and replacing them on the plane formed by the beam (22), the relationship between L _i and | Δ _i '|
It can be seen that the locus of _i ', y _i ') is constrained by the condition that it becomes a straight line (see FIG. 9).

Therefore, the locus of (x _i ', y _i ') is approximated to the following straight line by the method of least squares.

Y = a _y · x + b _y The intersection of the above straight line and y = x · tan θ _i (x _i ,
y _i ) is obtained, and is used to correct | Δ _i '|

| Δ _i '| = (x _i −x ₀ ) ² + (y _i −y ₀ ) ² (x ₀ , y ₀ ) = (L ₀ , 0) And finally, the corrected | Δ Δ _i ′ is corrected by the following equation using _i ′ │ and the slope a _z of the straight line (Z = a _z · x + b _z ).

[0070]

[Equation 20] Next, the operation of obtaining the posture matrix R of the reference coordinate system in step 1 of FIG. 13 will be described in detail with reference to the flowchart of FIG.

In FIG. 14, first, i is set to 1, and the jig (17) and the beam (22) are set to the initial posture (θ _x = 15 °,
θ _z = 0 °) (step 11). Next, the displacement vector Δ _i ′ (dx _i , 0, dz _i ) is measured (step 12). This will be described later. Then i and 7
(Step 13), and if i is smaller than 7, proceed to step 14, set (i + 1) to i, and set (θ _z −5 °) to θ _z to set the beam (22). Change the posture of and return to step 12. In step 13, i
If becomes 7 or more, the process proceeds to step 15 to perform regression correction of the displacement vector Δ _i ′ (dx _i , 0, dz _i ). This will also be described later. Next, the displacement vectors Δ ₁ 'and Δ ₂ ' at θ _z = ± 15 ° are extracted, and these are Δ ₁ ,
Set to Δ ₂ (step 16). Finally, each element of the posture matrix R of the reference coordinate system is calculated (step 17). This will also be described later.

Next, the operation of the displacement vector Δ _i 'measurement in step 12 of FIG. 14 will be described in detail with reference to the flowchart of FIG.

In FIG. 15, first, j is set to 1 (step 1201). Next, the output value of the beam movement detector (24) (the movement amount y _{j in} the Y-axis direction of the beam (22)) is read, and (y
_1- y _j ) is set to y _j (step 1202). Next, the coordinates (x _j , z _j ) of the slit end point are measured (step 1203). Next, j and 20 are compared (step 120
4) If j is smaller than 20, proceed to step 1205, set (j + 1) to j, and move the movable table (19).
That is, the beam (22) is moved in the Y-axis direction by a certain amount, and the step
Return to 1202. If j becomes 20 or more in step 1203, the flow advances to step 1206 to perform a linear regression calculation to obtain the following equation.

[0074]

[Equation 21] Then, finally, (a _x , 0, a _z ) is the displacement vector Δ
(Step 1207).

Next, the operation of the displacement vector regression correction in step 15 of FIG. 14 will be described in detail with reference to the flowcharts of FIGS. 16 and 17.

In FIGS. 16 and 17, first, the displacement vector Δ _i 'is subjected to regression correction to obtain the following equation (step 1501).

Z = a _z · x + b _z Next, i is set to 1 (step 1502) and the z component is corrected by the following equation (step 1503).

A _z · dx _i + b _z → dz _i Next, i and 7 are compared (step 1504). If i is smaller than 7, step 1505 is proceeded to, i + 1 is set to i, and step 1503 is entered. Return. I is 7 in step 1504
If so, the process proceeds to step 1506 to set i to 1. Then set Δ _i 'by the following equation (step
1507).

[0079] _{(dx i -dz 4, 0,} dz i -dz 4) → Δ i ' Next, the following equation to determine the theta _i, L _i (step 150
8).

[0080]

[Equation 22] Next, (x _i ', y _i ') is obtained by the following equation (step 1509).

L _i ((cos θ _i , sin θ _i ) → (x
_i ′, y _i ′) Next, i is compared with 7 (step 1510). If i is smaller than 7, the process proceeds to step 1511, i + 1 is set to i, and the process returns to step 1507. I is 7 in step 1510
If so, the process proceeds to step 1512, where (x _i ', y _i '
) Is performed to obtain the following equation.

Y = a _y · x + b _y Next, i is set to 1 (step 1513), and θ _i is obtained from the following equation (step 1514).

15 ° −5 ° · (i−1) → θ _i Next, the intersection (x _i , y) of the straight lines represented by the following two equations
_i ) is obtained (step 1515).

Y = a _y · x + b _y y = x · tan θ _i Next, i is compared with 7 (step 1516). If i is smaller than 7, the process proceeds to step 1517 to set i + 1 to i, Return to step 1514. I is 7 in step 1516
If so, the process proceeds to step 1518 and i is set to 1. Next, Δ _i 'is corrected by the following equation (step 15
19).

[0085]

[Equation 23] Next, Δ _i 'is corrected by the following equation (step 1520).

[0086]

[Equation 24] Next, i is compared with 7 (step 1521). If i is smaller than 7, the process proceeds to step 1522, i + 1 is set to i, and the process returns to step 1519. I is 7 in step 1521
When the above is reached, the processing is terminated.

Next, with reference to the flow chart of FIG. 18, the operation of calculating each element of the posture matrix R of the reference coordinate system in step 17 of FIG. 14 will be described in detail.

In FIG. 18, first, i is set to 1 (step 1701). Next, from Δ _i to (Δx
_i , 0, Δz _i ) is calculated (step 1702).

Δ _i → (Δx _i , 0, Δz _i ) Next, θ _i is calculated by the following equation (step 1703).

[0090] 15 ° -30 ° · (i- 1) → θ i Next, determine the L _i by the following equation (step 1704).

[0091]

[Equation 25] Next, (x _i , y _i , 0) is calculated by the following equation (step 1705).

(L _i · Rot (θ _i ) _z −E) · e _y
→ (x _i , y _i , 0) Next, i is compared with 2 (step 1706). If i is smaller than 2, the process proceeds to step 1707, i + 1 is set to i, and the process returns to step 1702. I is 2 in step 1706
If above, the process proceeds to step 1708, obtaining the n _x by the following equation (step 1708).

[0093]

[Equation 26] Next, ly is _calculated by the following equation (step 1709).

[0094]

[Equation 27] Next, _ny is calculated by the following equation (step 1710).

[0095]

[Equation 28] Next, φ is calculated by the following equation (step 1711).

Sin ⁻¹ n _y → φ Next, ω is obtained by the following equation (step 1712).

Sin ⁻¹ (−n _x / cos φ) → ω Next, ψ is obtained by the following equation (step 1713).

[0098] _{^{sin -1 (-l y · tanφ /}} n y) → ψ Then, a R a last equation (step 171
Four).

Rot (ψ) _z · Rot (φ) _x · Rot
(Ω) _y → R Next, with reference to FIG. 10, a method of calculating the traveling direction vector e _l of the line in step 2 of FIG. 13 will be described.

The beam (22) in the standard posture as shown in FIG.
When moving the base (18) in the direction of travel of the line,
That is, the displacement vector observed when the entire jig (17) (entire reference coordinate system) is moved in the traveling direction of the line is Δ
_{Given 0} , this would be:

[0101]

[Equation 29] Further, assuming that the moving amount of the line at this time is L '(O'P'), it is understood from FIG. 10 that Δ ₀ is expressed as follows.

Δ ₀ = L ₀ · _ey ′ −L ′ · e ₁ ( _ey ′: direction vector of beam in reference posture, L ₀ : O ′ P
₀ ) Therefore, the following equation is obtained.

E _l = (L ₀ · e _y '−Δ ₀ ) / L' (10) On the other hand, e _y 'is the direction vector of the Y axis (reference coordinate system), that is, the Y component of R. be equivalent to. Therefore, e _y 'is represented by the following equation.

[0104]

[Equation 30] In addition, _{here, L '= | L 0 ·} e y' -Δ 0 | a since, holds the following equation.

[0105] _{^{L '2 = (L 0 ·}} l y -Δx 0) 2 + (L
^{_{0 · m y) 2 + (}} L 0 · n y -Δz 0) 2 by solving this, L ₀ is calculated by the following equation (11).

[0106]

[Equation 31] Therefore, if the displacement vector Δ ₀ of the reference posture is measured, the direction vector e _l of the line in the head coordinate system is determined by the equations (10) and (11).

Next, referring to the flow chart of FIG. 19, the traveling direction vector e of the line in step 2 of FIG.
_{l The} operation of the operation will be described in detail.

In FIG. 19, first, the jig (17) and the beam
(22) is set to the reference posture (step 21), and the displacement vector Δ ₀ is measured (step 22). This will be described later. Next, e _y 'is calculated by the following equation (step 2
3).

R · e _y → e _y 'Next, L ₀ is calculated by the following equation (step 24).

[0110]

[Equation 32] Finally, e _l is calculated by the following equation (step 2
Five).

L ₀ · e _y ′ −Δ ₀ → e _l Next, the displacement vector measurement operation of step 22 of FIG. 19 will be described in detail with reference to the flowchart of FIG.

In FIG. 20, first, i is set to 1 (step 2201). Next, the output of the line movement amount detector (16) (Y _l axis direction movement amount y _{i of the} entire jig (17)) is read and (y ₁ −y _i ) is set to y _i (step 22).
02). Next, the coordinates (x _i , z _i ) of the slit end point are measured (step 2203). Next, i is compared with 20 (step 2204). If i is smaller than 20, the process proceeds to step 2205, where (i + 1) is set to i and the jig (1
7) The entire unit is moved in the Y ₁ -axis direction by a certain amount, and the process returns to step 2202. If i becomes 20 or more in step 2204, the flow advances to step 2206 to perform a linear regression calculation to obtain the following equation.

[0113]

[Expression 33] Then, finally, (a _x , 0, a _z ) is the displacement vector Δ
(Step 2207).

Next, how to obtain the correction coefficient in step 3 of FIG. 13 will be described.

Similarly to the above, from FIG. 10, the displacement vector Δ obtained when observing the beam (22) having an arbitrary posture (direction vector: e ₀ ) in the line coordinate system is expressed by the following equation.

Δ = L · e ₀ · −L ′ · e _l (L: O ′ P) Therefore, the following expression (12) is obtained.

[0117] _{e 0 = (L '· e} l + Δ) / | L' · e l + Δ | ····· (12) ( described in the measurement coordinate system) which is the deviation of the coordinate system e _l
Is a correction formula in consideration of, and e ₀ is the orientation vector itself stored in the data processing device (6). However, when the displacement vector is measured, the movement amount L = 1 is set for the sake of convenience.
The movement amount L ₀ of the beam (22) and the movement amount L ′ of the line are both 1.

Next, referring to FIG. 11, the above measuring device
The measurement principle according to (3) will be explained.

This shape measurement corrects the measurement direction vector, that is, the orientation vector e ₀ and the approach vector e _a at the time of actual measurement, using the line traveling direction vector e _l .

DUT measured with two heads (4) and (5)
From the slit end point coordinates P _1i (x _1i , y _1i , z _1i ) and P _2i (x _2i , y _2i , z _2i ) (i = 1 to n) in (2), the following regression linear equation (13), Get (14).

[0121]

[Equation 34] Here, y _1i and y _2i are both reading values of the line movement amount detector (16) of the conveyor (1), and for convenience, y ₁₁ = y ₂₁ = 0 at the measurement starting points P ₁₁ and P ₂₁ .

Also, in the coordinate system shown in FIG. 11, it should be noted that the Y coordinate value of the first head (4) has a negative direction in which the output value of the line movement amount detector (16) increases. I won't.

Next, using the slopes a _x1 and a _z1 of the straight line obtained above, the displacement vector Δ is expressed as follows.

Δ = (a _x1 , 0, a _z1 ) Then, e ₀ is calculated by the equation (12). However, at this time, L '= 1.

Two heads calculated in advance as described above
(4) Based on the height direction correction amount dz of (5), the distance W _off between the origins, and the straight line intercepts b _x1 , b _z1 , b _x2 , and b _z2 obtained above, the head of the first head (4) The measurement starting points P ₁₁ and P ₂₁ in the coordinate system are expressed as follows.

[0126] _{P 11 = (b x1, 0} , b z1) P 21 = (W off -b x2, 0, b z2 -dz) here, put the e _x 'as shown in the following formula.

[0127]

[Equation 35] Then, the following equation holds.

E _a · e ₀ = 0 e _a · e _x '= 0 However, | e _a | ² = 1. By solving this, e _a is calculated.

Next, the above-described shape measurement processing will be described with reference to the flowchart of FIG.

In FIG. 21, first, in step 201, the process waits until the measurement is started, and when the measurement is started, the first
Displacement vector (Δ ₁ , b _x1 , b _z1 ) by the head (4)
Is measured (step 202), and at the same time, the displacement vector (Δ ₂ , b _x2 , b _z2 ) is measured by the second head (5) (step 203). Next, z is calculated by the following equation (step 204).

B _z2 −b _z1 + dz → z Next, W is calculated by the following equation (step 205).

W _off − (b _x2 + b _x1 ) → W Next, e ′ _x is obtained by the following equation (step 206).

[0133]

[Equation 36] Next, e ₀ is corrected by the following equation (step 207).

(E ₁ + Δ ₁ ) / (| e ₁ + Δ ₁ |) → e ₀ Next, the following three equations are solved to correct e _a (step 208).

E _a · e ′ _x = 0 e _a · e ₀ = 0 | e _a | ² = 1 And finally, from the orientation vector e ₀ to the origin P _0.
Is calculated (step 209).

[0136]

As described above, according to the three-dimensional shape measuring method of the present invention, even if there is an angle deviation between the transport line and the measuring head coordinate system, a measurement error does not occur and the measured object can be measured. It is possible to accurately measure the three-dimensional shape of an object, and therefore, it is not necessary to strictly adjust the position of the measuring head, and the position adjustment is easy.

[Brief description of drawings]

FIG. 1 is a perspective view of a main part of a three-dimensional shape measuring apparatus showing an embodiment of the present invention.

FIG. 2 is a block diagram of a main part of a three-dimensional shape measuring apparatus.

FIG. 3 is a perspective view of a main part of a correction jig and a three-dimensional shape measuring apparatus.

FIG. 4 is an explanatory diagram showing an example of an image of a television camera.

FIG. 5 is an explanatory diagram for explaining the reason why a measurement error occurs due to a deviation in angle between the transport line and the measuring head coordinate system.

FIG. 6 is an explanatory diagram illustrating a method of obtaining a posture matrix of a reference coordinate system.

FIG. 7 is another explanatory diagram illustrating a method of obtaining a posture matrix of a reference coordinate system.

FIG. 8 is still another explanatory diagram illustrating a method of obtaining a posture matrix of a reference coordinate system.

FIG. 9 is still another explanatory diagram illustrating a method of obtaining a posture matrix of a reference coordinate system.

FIG. 10 is an explanatory diagram illustrating a method of calculating a traveling direction vector of a line.

FIG. 11 is an explanatory diagram illustrating a measurement principle of shape measurement.

FIG. 12 is a flowchart showing a method of adjusting the position of the measuring head and correcting the measuring head coordinate system.

FIG. 13 is a flowchart showing work for obtaining a correction formula.

FIG. 14 is a flowchart showing an operation of obtaining a posture matrix of a reference coordinate system.

FIG. 15 is a flowchart showing an operation of displacement vector measurement.

FIG. 16 is a flowchart showing the front part of the operation of displacement vector regression correction.

FIG. 17 is a flowchart showing a rear part of the operation of displacement vector regression correction.

FIG. 18 is a flowchart showing the operation of calculating each element of the attitude matrix of the reference coordinate system.

FIG. 19 is a flowchart showing the operation of calculating a traveling direction vector of a line.

FIG. 20 is a flowchart showing an operation of displacement vector measurement.

FIG. 21 is a flowchart showing a shape measurement process.

[Explanation of symbols]

 (1) Conveyor (conveying means) (2) Object to be measured (3) Three-dimensional shape measuring device (4) (5) Measuring head (6) Data processing device (7) (11) Slit light source (8) (12) TV camera (two-dimensional imaging device) (9) (13) Slit light (9a) (13a) Slit light surface (10) (14) Light cutting line (15) Conveyor drive device (16) Line movement amount detector (17) ) Correction jig (18) Base (19) Moving base (20) (21) Swiveling base (22) Beam (member to be measured) (23) Moving base drive (24) Beam movement detector (25) ( 27) Revolving platform drive (26) (28) Swing angle detector (29) Light cutting line (30) Light cutting line image (30a) Slit end point (A) Transfer line

Claims (2)

[Claims] 1. A measuring head provided with a slit light source and a two-dimensional imaging device is provided at a predetermined position around a conveyance line,
The object to be measured is moved along the carrying line by the carrying means, and at a plurality of object moving positions, the slit light source irradiates the object to be measured with slit light, and the slit light is formed by hitting the surface of the object to be measured. The optical cutting line is imaged by the two-dimensional imaging device, and the two-dimensional image data of the optical cutting line imaged by the two-dimensional imaging device is converted into three-dimensional position data in the three-dimensional measuring head coordinate system of the measuring head using the conversion position data. Then, in a three-dimensional shape measuring method for synthesizing three-dimensional position data at a plurality of moving positions of the measured object to obtain a three-dimensional shape of a required portion of the measured object, a dedicated correction jig is used, The position of the predetermined part of the jig in the measurement head coordinate system is measured using the measurement head in a plurality of predetermined postures, and the measurement result is processed into data to determine the transfer line and the measurement line. A deviation of an angle with respect to one coordinate axis of the head coordinate system and a correction expression for correcting this deviation are obtained, and the three-dimensional position data at each moving position of the object to be measured is corrected using this correction expression. A three-dimensional shape measuring method. 2. The correction jig comprises a base fixed to the transfer means and moved along the transfer line, and a bar-shaped member to be measured which extends substantially parallel to the transfer line in the reference posture. The correction jig is moved with respect to the base in the length direction of the reference posture by the moving means, and is rotated by the rotating means about two axes orthogonal to the moving direction. With the base fixed to the transfer means and the transfer means stopped, the member to be measured is moved with respect to the base by the moving means, and the slit light source causes slit light to reach the member to be measured at a plurality of positions to be measured. And the slit light strikes the surface of the member to be measured to form a light-section line, which is imaged by the two-dimensional image pickup device, and the secondary line of the light-section line imaged by the two-dimensional image pickup device. Based on the image data, the posture of the member to be measured in the reference posture with respect to one coordinate axis of the measuring head coordinate system is obtained, and the fixing jig is fixed by the conveying means with the member to be measured fixed to the reference posture with respect to the base. The two-dimensional image pickup device is moved along a transport line, and a slit light source irradiates a member to be measured with slit light at a plurality of jig moving positions, and the slit light strikes the surface of the member to be measured to form a light cutting line. Based on the two-dimensional image data of the optical cutting line imaged by the two-dimensional image pickup device and the posture of the member to be measured in the reference posture with respect to one coordinate axis of the measurement head coordinate system obtained previously, the conveyance line and the measurement head 3. The three-dimensional shape measuring method according to claim 1, wherein an angle deviation from one coordinate axis of the coordinate system and a correction formula are obtained.