JP6102829B2 - Absolute coordinate position measurement method by stereo matching method and absolute coordinate position measurement device by stereo matching method - Google Patents

Description

  This invention is a repetition of a similar shape of an object, and measurement is possible even when it is difficult to associate feature points with each other or when there is an obstacle between the imaging means and the object The present invention relates to an absolute coordinate position measurement method using a stereo matching method and an absolute coordinate position measurement device using a stereo matching method.

  A stereo matching method is known as a method for recognizing a three-dimensional shape. In the stereo matching method, three-dimensional information is obtained from images taken by two right and left photographing means. That is, an image obtained by performing a triangulation using a triangle composed of the center of two photographing means and a measuring point, and corresponding points of each point in an image photographed by one photographing means, is obtained by photographing the other photographing means. This is a method for performing the three-dimensional measurement of the target determined in the above.

  In such a stereo matching method, it is difficult to associate two images taken by the left and right photographing means. For example, the association is performed based on feature points in the image such as corners of three sides. Specifically, an image is cut out as a block-shaped partial image, a portion having the highest matching rate is selected, and feature points are associated with each other. However, if the target object has a similar shape (for example, in the case of power equipment such as a terminal board), the extracted partial images are all similar, and it is difficult to extract feature points. Become.

  In addition, since the left and right photographing means are fixed, measurement cannot be performed when there is an obstacle between the photographing means and the object.

  A stereo matching processing system, a stereo matching processing method, and a program for accurately associating the same position between a plurality of images are known (see, for example, Patent Document 1). In this technique, in a plurality of images obtained by photographing the same object from different directions, regions having the maximum correlation coefficient on the same scanning line are associated with each other as indicating the same position. And it is discriminate | determined whether the line segment matched as what shows the same position to each of several images was drawn. Then, when it is determined that the line segment is drawn, the intersections of the scan line and the line segment are associated with each other as indicating the same position, instead of the regions having the maximum correlation coefficient on the same scan line. is there.

JP 2010-128608 A

  In the apparatus of Patent Document 1, when the object has a similar shape, it is difficult to extract feature points. In addition, measurement is not possible when there is an obstacle between the photographing means and the object.

  Therefore, the present invention measures the object even if the object has a similar shape and it is difficult to associate the feature points with each other or when there is an obstacle between the imaging means and the object. It is an object to provide an absolute coordinate position measurement method using a stereo matching method and an absolute coordinate position measurement device using a stereo matching method.

In order to solve the above-mentioned problem, the invention of claim 1 is a first imaging means for imaging an object, and a first imaging means for imaging the object that is movable in a predetermined direction with respect to the first imaging means. 2 is an absolute coordinate position measurement method using a stereo matching method using the imaging unit, and is positioned in front of the object and a straight line connecting the first imaging unit and the second imaging unit. When the straight line connecting the second imaging means and the object is orthogonal, the second imaging means is positioned in front of the object, and the first imaging means satisfies a predetermined condition As a reference point, θ ₁ is the incident angle of the first imaging unit, θ ₂ is the incident angle of the second imaging unit, and d is the first imaging unit after moving. The distance from the second imaging means, l, the absolute coordinates of the object and the first When the direction d of the distance between the absolute coordinate position of the imaging means, and, tan .theta _{1 /} tan .theta _{2 =} by moving the second imaging means so as to satisfy the -d / l + 1 conditions, the first An absolute coordinate position measurement method based on a stereo matching method, characterized in that the position of the object is calculated based on image information taken by a photographing means, focal length, and l.

  According to the present invention, with the second imaging means positioned in front of the object, the first imaging means is positioned so as to satisfy the predetermined condition as a base point, and the second imaging means is moved. The position of the object is measured based on the image information photographed by the second photographing means.

According to a second aspect of the present invention, there is provided a first photographing means for photographing an object, a second photographing means for photographing the object, which is movable in a predetermined direction with respect to the first photographing means, and the first In the case where a straight line connecting the first imaging unit and the second imaging unit and a straight line connecting the second imaging unit and the object when positioned in front of the object are orthogonal to each other, the first With the two imaging means positioned in front of the object, the first imaging means is positioned so as to satisfy a predetermined condition as a base point, and θ ₁ is an incident angle of the first imaging means, θ ₂ is the incident angle of the second imaging means, d is the distance between the first imaging means and the second imaging means after movement, l is the absolute coordinates of the object and the first of the d direction distance between the absolute coordinate position of the imaging unit, when a, tanθ ₁ / tanθ ₂ = - The second imaging unit is moved so as to satisfy the condition of / l + 1, and the position of the object is calculated based on the image information captured by the first imaging unit, the focal length, and the l. And an absolute coordinate position measuring device using a stereo matching method.

  According to a third aspect of the present invention, in the second aspect of the present invention, the apparatus includes a moving unit having a guide unit that moves the second photographing unit in a predetermined direction with respect to the first photographing unit. And

  According to the first and second aspects of the present invention, when the first photographing means is positioned so as to satisfy the predetermined condition as a base point and the second photographing means is moved, the image information photographed by the second photographing means is displayed. Based on this, the position of the object can be measured by the stereo matching method. For this reason, even if the object has a similar shape and it is difficult to associate feature points with each other, or even if there is an obstacle between the imaging means and the object, the position of the object Can be easily measured by the stereo matching method.

  According to the invention of claim 3, since the second photographing means can move the second photographing means in a predetermined direction with respect to the first photographing means by the moving means, it is possible to more easily detect the object. The position can be measured by a stereo matching method.

It is a schematic block diagram which shows the absolute coordinate position measuring apparatus by the stereo matching method in embodiment of this invention. It is the schematic which shows the positional relationship of each part in the apparatus of FIG. It is the schematic which shows the positional relationship of each part in the apparatus of FIG. It is the schematic of the structure which measures the position of the 1st imaging | photography means and the 2nd imaging | photography means in the apparatus of FIG. It is the schematic which shows the apparatus and measuring object of FIG. It is a flowchart which shows the information processing which calculates the absolute coordinate position in the apparatus of FIG.

  The present invention will be described below based on the illustrated embodiments.

1 and 2 are schematic configuration diagrams showing an absolute coordinate position measuring apparatus (hereinafter referred to as “absolute coordinate position measuring apparatus”) 1 by a stereo matching method according to an embodiment of the present invention. Absolute coordinate position measuring apparatus 1 mainly includes a camera (left) O _L as the first imaging means, and the camera (right) O _R as a second photographing means, a moving jig 2 as a moving means, the absolute And a calculation device 3 (not shown) as coordinate position calculation means. In this embodiment, the absolute coordinates (X _p , Y _p , Z _p ) of the measurement object P are calculated.

Camera (left) O _L is for photographing a measuring object P.

Camera (right) O _R is movable in a predetermined direction (straight) to the camera (left) O _L, is to shoot the measurement object P. This camera (left) O _L and the camera (right) O _R, photographing the same measuring object P from different directions.

Moving jig 2, the camera has a guide portion for moving in a predetermined direction (right) _{O R} to the camera (left) _{O L,} the camera (left) and _{O L,} the position of the camera (right) _{O R} It is a jig for setting. The moving jig 2 mainly includes a first rail 21, a second rail 22, a first arm 23, and a second arm 24 as guide portions.

  The first rail 21 is fixed to a wall surface or the like so as to be parallel to the measurement object P in the X-axis direction.

  The second rail 22 is fixed to a wall surface or the like so as to be parallel to the first rail 21 at an interval m.

The first arm 23 includes a first rail 21 and are slidably disposed in the X-axis direction along the second rail 22, the camera (left) O _L to movably support the X-axis direction Is. The first arm 23 is formed in a substantially U shape by a first base portion 23a, a second base portion 23b, and a connecting portion 23c. The first base portion 23a is formed with an insertion hole 23d through which the first rail 21 is inserted, and the second base portion 23b is formed with an insertion hole 23e through which the first rail 21 is inserted. The connecting portion 23c connects the first base portion 23a and the second base portion 23b.

The second arm 24 is formed in the same manner as the first arm 23, and is slidable in the X-axis direction along the first rail 21 and the second rail 22. ) and O _R is for movably supporting the X-axis direction.

  The infrared light emitting portion S is disposed on the first base portion 24a of the second arm 24, and emits infrared rays toward the infrared light receiving portions C1 and C2.

The infrared light receivers C1 and C2 measure the reception angles θ ₁ and θ _{2 at} which the received level of the received infrared light is maximum, and transmit them to the calculation device 3 by wire or wirelessly. The infrared light receiving portions C1 and C2 are disposed opposite to the second base portion 23b of the first arm 23 with the insertion hole 23e interposed therebetween. As shown in FIG. 3, when the distance between the infrared light receiving portions C1 and C2 is Q, the infrared light receiving portions C1 and C2 are disposed at positions separated from each other by a distance Q / 2 from the insertion hole 23e.

Here, the camera (left) _{O L} and the camera (right) method of calculating the distance x between _{O R} will be described. The distance x is derived from the distance Q and the angles θ ₁ and θ _{2 as} follows.
_{tanθ 1 = (m + Q /} 2) by transforming the / x, m + Q / 2 = xtanθ 1 ·· (1)
_{tanθ 2 = (m-Q /} 2) by modifying the / x, m-Q / 2 = xtanθ 2 ·· (2)
From (1) + (2), Q = x (tan θ ₁ −tan θ ₂ ) (3)
Furthermore, (3) is transformed,
x = Q / (tan θ ₁ −tan θ ₂ ) (4)
Get.

Thus calculated, the camera (left) O _L and the camera (right) distance x between O _R also moves to any position of two cameras, the angle theta _1, from theta ₂ (4) equation it can. That is, the angle theta _1, from theta _2, the movement distance of the camera (right) _{O R} from the camera (left) _{O L} was the starting point is calculated.

In the moving jig 2 having such a configuration, as shown in FIG. 4, an optical fiber cable F extending from the light source A is disposed along the second rail 22, and below the optical fiber cable F, A total reflection layer T that reflects the upper light is disposed. Then, when the first arm 23 and the second arm 24 move on the second rail 22, the detection count number of light leaking from the light output hole H is acquired. And the absolute value of the horizontal position of the 1st arm 23 or the 2nd arm 24 is acquired based on the detection count number of the detected light. Thus, to obtain the camera (left) and _{O L,} the position of the camera (right) _{O R.}

Next, an absolute coordinate position measurement method using the stereo matching method using the absolute coordinate position measurement apparatus 1 will be described. Here, as shown in FIG. 5, the camera (left) _{O L} incident angle theta ₁ of the camera (Right) _{O R} incident angle θ ₀ = 90 °, the post-movement between the focal length f, the camera (left) _{O L} camera (right) _O distance between _R d, the absolute coordinates of the measurement object _{P (X p, Y p,} Z p) d of the camera (left) absolute coordinate position of the _{O L} (0,0,0) Let the distance in the direction be l.

First, the camera (right) _{O R} a measurement object _{P (X P, Y P,} Z P) is set in front of the d direction and d = l. At this time, an image camera (right) O _R is taken may be front view.

Next, the camera (left) _{O L} incident angle theta ₁ in a fixed state of (while fixing the camera (left) _{O L),} a camera (right) _{O R,} linearly in the direction d (first Along the rail 21 and the second rail 22 in the X-axis direction) to the position of the incident angle θ ₂ . Camera (right) O _R selects the position where no operator M between the measurement object P and the camera (right) O _R.

In this _{case, y = X R / X L} , as x = d, y = if -x + 1 relationship is established, by Lee type, based only on the information of the camera (left) _{O L,} the measurement object P ( X _P , Y _P , Z _P ) can be calculated.

Derivation of the equation (a) will be described. The principle of the stereo matching method is the following equation.

When the formula is transformed, the following formula is obtained.

Camera (left) _{O L,} the camera (right) _O 2 triangles _S L of _R, Focusing on _{S R,} the following two equations are obtained.
X _R / X _L = (f / tan θ ₂ ) / (f / tan θ ₁ )
= Tan θ ₁ / tan θ ₂ ...

First, the camera such that the incident angle θ ₀ = 90 ° (right) _{O R,} sets the camera (right) _{O R} measurement object _{P (X P, Y P,} Z P) in front in the direction d of D = l. Thus, for example, when the camera (left) O _L to the base point (fixed) as the incident angle θ _{1 =} 30 °, the next host expression is obtained from B-type.

  Therefore, the formula E is the same as the formula A.

Next, the camera angle of incidence theta ₁ (left) _{O L} in a state where a base point (fixed to θ ₁ = 30 °), the camera (right) _{O R} the incident angle θ ₂ = 60 ° in a straight line in direction d Or the case where it moves to the position of θ ₂ = 75 ° will be described.
(1) Camera (right) when you move the _{O R} to the position of the incident angle _{θ 2 = 60 ° y = X} R / X L, as x = d, y = -x + 1 (x = 1-y) of the equation Using,
y ₁ = tan (30 °) / tan (60 °)
= 0.5774 ÷ 1.7321 ≒ 0.3334
∴x = 1−y ₁ = 1−0.3334 = 0.6666
That is, if set camera (right) _{O R} in the position of 0.6666L, the measurement object _{P (X P, Y P,} Z P) to be calculated by the Lee type.

Check the result. It is calculated as follows from the formula (c).
X _P = d / (1-X _R / X _L ) = 0.6666 l / (1-0.3334) = l
Y _P = (d · Y _L / X _L ) / (1-X _R / X _L )
= 0.6666 l · (Y _L / X _L ) / (1-0.3334) = (Y _L / X _L ) · l
Z _P = ((d · f) / X _L )) / (1-X _R / X _L )
= ((0.6666l · f) / X _L ) / (1-0.3334) = (f · l) / X _L

(2) Similar to the camera when you move (right) _{O R} to the position of the incident angle _{θ 2 = 75 ° (1)} ,
y ₂ = tan (30 °) / tan (75 °)
= 0.5774 ÷ 3.7321 ≒ 0.1547
∴x = 1−y ₂ = 1−0.1547 = 0.8453
That is, if set camera (right) _{O R} in the position of 0.8453L, the measurement object _{P (X P, Y P,} Z P) to be calculated by the Lee type.

Check the result. It is calculated as follows from the formula (c).
_{X P = d / (1-} X R / X L) = 0.8453l / (1-0.1547) = l
Y _P = (d · Y _L / X _L ) / (1-X _R / X _L )
= 0.8453 l · (Y _L / X _L ) / (1-0.1547) = (Y _L / X _L ) · l
Z _P = ((d · f) / X _L )) / (1-X _R / X _L )
= ((0.8453l · f) / X _L ) / (1-0.1547) = (f · l) / X _L

Thus, the moving position of the camera (right) O _R, w, is calculated so as to satisfy the condition based on the value such as tan .theta _2, by setting the camera (right) O _R in position, Lee formula Accordingly, only the absolute coordinate position of the measurement object P image information of the camera _{(left) O L (X P, Y} P, Z P) can be calculated. Incidentally, l, f is a constant value, w is the distance to P in the Z axis direction, the theta ₂ is the incident angle of the camera (right) O _R.

That, w calculated in this manner, so as to satisfy the like tan .theta _2, may be set the moving position of the camera (right) O _R.

  It continues and demonstrates using a specific numerical value.

Since the camera (left) incident angle theta ₁ of the _{O L} is 30 °, from two equations,
y = X _R / X _L = tan (30 °) / tan θ ₂ = 0.5774 / tan θ ₂
Using the relational expression y = −x + 1 (x = 1−y),
x = 1−0.5774 / tan θ ₂ ... Formula In the triangle PO _R B shown in FIG.
tan θ ₂ = w / (l−d) ...... G equation where l-d = 0.5l and w = l
tan θ ₂ = w / (l−d) = l / 0.5 l = 2 (θ ₂ is about 64 °)
Substituting θ ₂ into the formula,
x = 1-0.5774 / 2 = 0.7113
Therefore, d = 0.7113l is set.

Further, y = X _R / X _L
= Tan (30 °) / tan θ ₂ = 0.5774 / 2 = 0.2887
From formula C,
_{X P = d / (1-} X R / X L) = 0.7113l / (1-0.2887) = l
Y _P = (d · Y _L / X _L ) / (1-X _R / X _L )
= 0.7113 l · (Y _L / X _L ) / (1−0.2887) = (Y _L / X _L ) · l
Z _P = (d · f · 1 / X _L ) / (1-X _R / X _L )
= ((0.7113l · f) / X _L )) / (1-0.2887) = (f / X _L ) · l

Accordingly, if, l = 1 m, as w = 1 m, if the set camera (left) _{O L} at the position of d = 0.7113m, the absolute coordinate position of the measurement object P by Lee formula _{(X P, Y} P, Z _P ) can be calculated.

  Further, the following two cases will be described.

<Case 1> When l−d = 0.2 l, w = l
tan θ ₂ = w / (1-d) = l / 0.2l = 5 (θ ₂ is about 79 °)
Substituting θ ₂ into the formula,
x = 1-0.5774 / 5 = 0.8845
Therefore, d = 0.8845 l is set.

Further, y = X _R / X _L = tan (30 °) / tan θ ₂ = 0.5774 / 5 = 0.1155
From formula C,
X _P = d / (1-X _R / X _L ) = 0.8845 l / (1-0.1155) = l
Y _P = (d · Y _L / X _L ) / (1-X _R / X _L )
= 0.8845 l · (Y _L / X _L ) / (1-0.1155) = (Y _L / X _L ) · l
Z _P = (d · f · 1 / X _L ) / (1-X _R / X _L )
= ((0.8845L · f) / X _L )) / (1-0.1155) = (f / X _L ) · l
If w = l = 1 m, then d = (1−0.2) l = 0.8l = 0.8m

<Case 2> When l−d = 0.5 l, w = 3 l
tan θ ₂ = w / (1-d) = 3 l / 0.5 l = 6 (θ ₂ is about 80.5 °)
Substituting θ ₂ into the formula,
x = 1-0.5774 / 6 = 0.904
Therefore, d = 0.9041 is set.

Also, y = X _R / X _L = tan (30 °) / tan θ ₂ = 0.5774 / 6 = 0.096
From formula C,
X _P = d / (1-X _R / X _L ) = 0.904 l / (1-0.096) = l
Y _P = (d · Y _L / X _L ) / (1-X _R / X _L )
= 0.904 l · (Y _L / X _L ) / (1-0.096) = (Y _L / X _L ) · l
Z _P = (d · f · 1 / X _L ) / (1-X _R / X _L )
= ((0.904l · f) / X _L )) / (1-0.096) = (f / XL) · l
If l = 1m and w = 3m, then d = (1-0.5) l = 0.5l = 0.5m

As a prerequisite, the camera (left) _{O L} incident angle 30 °, and when f is a constant value, in the case of the case 1, the camera (right) _{O R} is w = l = 1 m, the d = 0.8 m Set it to the position. Further, in Case 2, the camera (right) _{O R} is, l = 1m, w = 3m , may be set at the position of d = 0.5 m.

Thus, l, w, d, be selected so that the value of X R _/ X _L satisfies certain conditions, camera (right) O _R moves linearly, the distance d between the camera is changed can also be a b-type to calculate the absolute coordinate position of the measurement object P only image information of the camera (left) O _L.

Accordingly, the appropriate conditions (y = -x + 1), by selecting some of the candidates for the position of the camera (right) O _R when moving to avoid the operator, the absolute coordinates of the measurement object P by Lee formula (X _P , Y _P , Z _P ) can be calculated quickly, easily and accurately.

As described above, the absolute coordinate position measuring method according to the stereo matching method according to the absolute coordinate position measuring apparatus 1, a camera (left) O _L is a base point is positioned at a predetermined condition is satisfied, the camera (right) O When _R moves to the predetermined condition is satisfied, the absolute coordinate position of the measurement object P can be measured by a stereo matching method based upon the image information captured by the camera (left) O _L. For this reason, even if the measurement object P is a repetition of a similar shape and it is difficult to associate feature points with each other or there is an obstacle between the photographing means and the measurement object, measurement is possible. The absolute coordinate position of the object P can be easily and accurately measured by the stereo matching method.

The measurement camera (right) O _R is the mobile jig 2, since the camera (left) O _L can be moved in a predetermined direction, the more easily the position of the measurement object P by the stereo matching method can do. Also, by moving the jig 2, the camera (left) _{O L,} the camera (right) _{O R} is freely movable, so camera (left) interval _{O L} and the camera (right) _{O R} can also be freely set, The measuring object P is not limited and can be used for various measuring objects P.

Thus, l, w, d, by selecting a value for X R _/ X _L in accordance with a predetermined condition, the camera (right) O _R moves linearly, even if the distance d between the camera is changed, Lee the equation can calculate the absolute coordinate position easily only image information of the camera (left) O _L.

Accordingly, the appropriate conditions (y = -x + 1), the candidate of the position of the camera (right) O _R when moving to avoid the operator, by preparing some measured object by Lee formula P The absolute coordinate position (X _P , Y _P , Z _P ) can be measured quickly, easily and accurately.

Although the embodiment of the present invention has been described above, the specific configuration is not limited to the above embodiment, and even if there is a design change or the like without departing from the gist of the present invention, Included in the invention. For example, a camera and a (left) _{O L,} the position of the camera (right) _{O R,} set automatically by the mobile jig 2, the absolute coordinate position of the measurement object _{P (X P, Y P,} Z P) is measured You may be made to do. Specifically, the control unit of the calculation device 3 is configured to include a program and a task for performing information processing along the flow shown in FIG. First, the control unit, the camera (right) O _R is set in front of the measurement object P (step S1). Then, the control unit, the camera (left) _{O L} is set to a base point (step S2). Then, the calculation device 3, the position of the camera (right) _{O R} is selected (step S3). Then, the control unit, the camera (right) _{O R} is moved to the selected position in step S3 (step S4). Then, the absolute coordinate position of the measurement object P is calculated by the calculation device 3 (step S5). In this manner, by the control unit, the camera and the (left) O _L, the position of the camera (right) O _R is set automatically by the mobile jig 2, the absolute coordinate position of the measurement object P is calculated.

Also, calculating device 3 calculated by the control unit, the camera (left) and the distance x between _{O L} and the camera (right) _{O R,} the absolute coordinate position of the measurement object _{P (X p, Y p,} Z p) a Of course, it is possible to use a program having a function to execute a function and a device capable of executing a task.

1 Absolute coordinate position measuring device (Absolute coordinate position measuring device using stereo matching method)
2 Moving jig (moving means)
21 First rail (guide section)
22 Second rail (guide section)
3. Calculation device (absolute coordinate position calculation means)
_OL camera (left) (first photographing means)
O _R camera (right) (second photographing means)
d Inter-camera distance f Focal length θ Incident angle S Infrared light emitter C1 Infrared light receiver C2 Infrared light receiver P Measurement object

Claims (3)

Absolute coordinates by a stereo matching method using a first imaging unit that images an object and a second imaging unit that can move in a predetermined direction with respect to the first imaging unit and that images the object. A position measurement method,
In a case where a straight line connecting the first imaging unit and the second imaging unit and a straight line connecting the second imaging unit and the object when positioned in front of the object are orthogonal,
Positioning the second imaging means in front of the object;
The first photographing means is positioned so as to satisfy a predetermined condition and serves as a base point.
θ ₁ is the incident angle of the first imaging unit, θ ₂ is the incident angle of the second imaging unit, and d is the distance between the first imaging unit and the moved second imaging unit. , L is the distance in the d direction between the absolute coordinate of the object and the absolute coordinate position of the first imaging means,
tan θ ₁ / tan θ ₂ = −d / l + 1
The second photographing means is moved so as to satisfy the condition of
Calculating the position of the object on the basis of the image information photographed by the first photographing means, the focal length, and the l;
The absolute coordinate position measuring method by the stereo matching method characterized by this. First imaging means for imaging an object;
A second photographing means that is movable in a predetermined direction with respect to the first photographing means and photographs the object;
In a case where a straight line connecting the first imaging unit and the second imaging unit and a straight line connecting the second imaging unit and the object when positioned in front of the object are orthogonal, With the second imaging means positioned in front of the object, the first imaging means is positioned so as to satisfy a predetermined condition as a base point, and θ ₁ is incident on the first imaging means. An angle, θ ₂ , an incident angle of the second imaging means, d, a distance between the first imaging means and the second imaging means after movement, l, an absolute coordinate of the object and the When the distance in the d direction with respect to the absolute coordinate position of the first imaging unit is set, the second imaging unit is moved so as to satisfy the condition of tan θ ₁ / tan θ ₂ = −d / l + 1, and Based on the image information photographed by the first photographing means, the focal length, and the l The absolute coordinate position calculation means for calculating the position of an object,
An absolute coordinate position measuring apparatus using a stereo matching method, comprising: A moving means having a guide part for moving the second photographing means in a predetermined direction with respect to the first photographing means;
The absolute coordinate position measuring apparatus by the stereo matching method according to claim 2, comprising:

Images