JP2718249B2 - Robot displacement detection device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for detecting a displacement of a robot,
The present invention relates to an apparatus capable of quantitatively detecting a positional shift and an angular shift with respect to a predetermined reference position.

2.Description of the Related ArtFor example, the paint surface inspection apparatus shown in FIG. 12 captures an image 10 ′ in which a luminous body 10 emitting scattered light having a stripe pattern is reflected on a paint surface 12 by a CCD camera 16 via a mirror 14, The smoothness and the like of the painted surface 12 are inspected from the width and interval of the striped pattern projected on the monitor 20 via 18, and the light emitting body 10, the mirror 14, and the CCD camera 16 are fixed in position on the robot hand 22. And is sequentially moved to another inspection position (teaching point) in accordance with the teaching position data stored in advance together with the robot hand 22 to inspect the painted surface 12. When teaching the teaching position data in such a coating inspection apparatus, a large number of teaching points T are set in advance on the coating surface 12 as shown in FIG. Using the teaching probe 24, the tip of the probe 24 is made to coincide with the teaching point T so that the teaching point T is located at the center of the imaging field of view of the CCD camera 16, and the probe 24 and the painted surface 12 are Is performed so that the formed angle β matches a predetermined design value.

By the way, the probe 24 has a painted surface during teaching.
Because it has flexibility so as not to scratch the 12, it is difficult to accurately position, and since the painted surface 12 has a complicated curved surface, it is also difficult to accurately measure the angle β, High-precision teaching was not always possible. For this reason, as shown in FIG. 15, when the center of the striped plate 26 attached to the light emitting body 10, that is, the robot hand 22 is correctly taught, the CCD
A mark M that can be distinguished from a defect existing on the teaching point T or the painting surface 10 by a color, shading, shape, or the like is provided at a portion located at the center of an image captured by the camera 16, and the robot hand 22 is driven according to the teaching position data. The position of the robot hand 22 is corrected from the position of the teaching point T on the image of the monitor 20, and the painted surface of the robot hand 22 is corrected from the position of the mark M on the image of the monitor 20.
It is conceivable to correct the angle shift with respect to 10.

Problems to be Solved by the Invention However, the positions of the teaching point T and the mark M on the image captured by the CCD camera do not change independently of each other due to the positional deviation and the angular deviation of the robot hand, but are related to each other. Therefore, in order to eliminate both the displacement of the teaching point T and the mark M on the image, for example, as shown in the flowchart of FIG. 16, the correction of the displacement of the robot hand and the correction of the angle displacement are performed. Must be repeated alternately by trial and error, which is troublesome and takes a long time, and the measurement of the amount of deviation is also performed by a worker using a ruler or the like on the monitor screen, so that high accuracy is not necessarily obtained. .

As a device for quantitatively detecting such a positional shift of the robot, for example, a device based on a light cutting method using slit light is disclosed in Japanese Patent Application Laid-Open No. 60-118907. In this case, Although positional displacement in the three-dimensional direction can be detected, it cannot be detected until angular displacement.

The present invention has been made in view of the above circumstances, and an object of the present invention is to determine a displacement dimension and a displacement angle of a movable portion such as a robot hand based on an image captured by an imaging device such as a CCD camera. An object of the present invention is to provide a position shift detecting device that can automatically detect the position shift.

Means for Solving the Problems In order to achieve the above object, not only the deviation of the teaching point T (first reference point) and the mark M (second reference point) on the image but also the dimensional error of the magnitudes thereof. In the present invention, the movable portion of the robot that is relatively moved with respect to the mirror-like reference surface that reflects light is moved relative to a first reference point provided on the reference surface. A device for detecting whether or not the camera is positioned at a predetermined reference position including an attitude with respect to the reference plane,
(A) a second reference point is provided at a predetermined position,
A second reference member fixedly disposed on the movable portion so that the second reference point is reflected on the reference surface; and (b) moving the movable portion to a position near the reference position,
An image pickup device fixedly disposed on a movable portion of the image pickup device so that the second reference point reflected on the reference plane and the first reference point enter the image pickup field of view; Between the first reference point on the displayed image and a predetermined first target position where the first reference point is located on the image when the movable part is located at the reference position. First displacement detection means for detecting dimensions, (d) the second reference point on the image,
A second displacement detection unit that detects a displacement dimension from a predetermined second target position where the second reference point is located on the image when the movable unit is located at the reference position; (E) the size of at least one of the first reference point and the second reference point on the image, and the size of one of the reference points on the image when the movable portion is located at the reference position Dimension error detecting means for detecting an error from a predetermined target dimension, and (f) the image detected by the first position error detecting means, second position error detecting means, and dimensional error detecting means, respectively. And a shift amount determining means for obtaining a shift size and a shift angle of the movable portion with respect to the reference position based on the position shift size and the size error of the size. To.

Functions and Effects of the Invention In such a device for detecting a position shift of a robot,
At least one of a displacement dimension between the first reference point and the first target position, a displacement dimension between the second reference point and the second target position, and a first reference point and a second reference point on an image captured by the imaging device. The size error between the size of the target and the target size is determined by the first position shift detecting unit, the second position shift detecting unit,
The displacement amount and the displacement angle with respect to the reference position are obtained by the displacement amount determining means based on the positional displacement dimension and the dimensional error.

As described above, according to the position shift detecting device of the present invention, the position shift of the movable portion with respect to the first reference point and the angle shift with respect to the reference plane are quantitatively detected. The teaching position data is corrected based on the above, so that the teaching position data is automatically corrected in a short time and at a higher rate than in the conventional case where the worker corrects the position shift and the angle shift by trial and error. It can be done with precision.

Embodiment Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the following embodiments, portions common to the above-described conventional example are denoted by the same reference numerals, and detailed description thereof will be omitted.

In FIG. 1, the robot hand 22 is three-dimensionally moved in directions of X, Y, and Z axes perpendicular to each other by the robot 30, and is rotated around the X axis and the Y axis to change its posture. It has become. robot
The teaching position data stored in the personal computer 32 in advance by the teaching process performed using the probe 24 and the like is loaded into the robot controller 34, and the robot hand 22 is moved to a plurality of teaching points in accordance with the teaching position data. It is controlled so as to sequentially move on T in a predetermined posture.

If the above teaching position data is accurate, the robot hand
Reference numeral 22 denotes a luminous body 10 'reflected on the teaching point T and the painted surface 12 at the center of the imaging field of the CCD camera 16 with respect to each teaching point T.
Are positioned at the reference position where the mark M is located, and the teaching point T and the images T ′ and M ′ of the mark M on the image P of the CCD camera 16 are respectively positioned at the image center O, but the X, Y, and Z axes If there is a positional deviation in the direction or an angular deviation about the X axis and the Y axis, the images T 'and M' are shifted from the center O of the image P as shown in FIG. Further, the mark M of the present embodiment has a square shape as shown in FIG. 3, and the size of the image M ′ on the image P depends on the positional deviation and the angular deviation of the robot hand 22. Will change accordingly. The mark M is provided on the striped plate 26 so that it can be distinguished from a defect existing on the teaching point T or the painted surface 10 by color, shading, shape, and the like.

In this embodiment, the painted surface 12 corresponds to a mirror-like reference surface that reflects light, the CCD camera 16 corresponds to an imaging device, and the robot hand 22 corresponds to a movable part of a robot, and is provided on the painted surface 12. Teaching point T corresponds to the first reference point,
The mark M provided in the striped plate corresponds to the second reference point.
26 corresponds to the second reference member, and the center O of the image P corresponds to the first target position and the second target position.

Then, an image signal representing the image P output from the CCD camera 16 is supplied to an image processing device 36. The image processing device 36 is configured to include a microcomputer,
The signal processing is performed according to the program stored in the ROM while utilizing the temporary storage function of the RAM. By performing the image processing according to the flowchart shown in FIG. The position deviation size and the deviation angle of the robot hand 22 are detected from the deviation size of the '′ and the size of the image M'.

In FIG. 4, at step S1, an image signal representing an image P is fetched from the CCD camera 16, and at step S2, an image T 'of the teaching point T is extracted from the image P. This extraction processing is performed by deleting the stripe pattern and the image M ′ from the image P by differentiation or the like. Then, in the next step S3, the deviation dimensions δX ₁ and δ between the center coordinates of the image T ′ and the image center O are determined.
Calculating a Y _1, it is stored in the memory 38 of the computer 32. These shift dimensions δX ₁ , δY ₁ represent the shift amount in the X-axis direction and the shift amount in the Y-axis direction, respectively, as is apparent from FIG. In the series of signal processing performed by the image processing device 36, the part that executes steps S2 and S3 corresponds to a first displacement detection unit.

In the following step S4, the image M 'of the mark M is extracted by removing the stripe pattern or the image T' from the image P by differentiation or the like. In the next step S5, the center coordinates of the image M 'and the image center O are compared. The shift dimensions ΔX ₂ and ΔY ₂ are calculated and stored in the memory 38 of the personal computer 32. These deviation dimensions δX ₂ ,
δY ₂ represents a shift amount in the X-axis direction and a shift amount in the Y-axis direction, respectively, as is apparent from FIG. In a series of signal processing by the image processing device 36, these steps S4
The step of executing steps S5 and S5 corresponds to a second displacement detection unit.

In the next step S6, the X-axis dimension a _X ′ and the Y-axis dimension a _Y ′ of the image M ′ are calculated, and a dimensional error δa _X from the preset target dimensions a _X and a _Y is calculated. (= A _X ′
−a _X ) and δa _Y (= a _Y ′ −a _Y ), and store the dimensional errors δa _X and δa _Y in the memory 38 of the personal computer 32. The target dimensions a _X and a _Y are previously stored in a ROM or the like as the dimensions of the image M ′ of the mark M on the image P when the robot hand 22 is accurately positioned at the reference position. Among a series of signal processing by the image processing device 36,
The part that executes step S6 corresponds to a dimensional error detecting unit.

Then, in step S7, the deviation size [delta] X ₁ stored in the memory _{38, δY 1, δX 2,} δY 2, dimensional error .delta.a _X, .delta.a _Y
And the displacement dimension ΔX in the X-axis direction and the displacement dimension Δ in the Y-axis direction of the robot hand 22 in accordance with a predetermined arithmetic expression.
Calculated Y, Z axis direction displacement dimension [Delta] Z, offset angle alpha _X around the X axis, the offset angle alpha _Y around the Y axis.

Here, the positional deviation and the angular deviation of the robot hand 22 will be specifically described. For example, as shown in FIG. 5, the light emitting body 10 and the CCD have a positional deviation of the dimension ΔX only in the X-axis direction. If the camera 16 is is positioned within the X-Z plane, the image T 'on the image P is the X-axis direction from the center O, as shown in Figure 6 [delta] X ₁
Just shift. The deviation dimension δX ₁ is as follows: θ is the viewing angle of the CCD camera 16, δθ ₁ is the deviation angle between the center of the field of view of the CCD camera 16 and the teaching point T, and P _X is the dimension of the image P in the X-axis direction. has a relationship represented by the following equation (1) between, theta and P _X displaced from it is known by the [delta] X ₁ angle .delta..theta ₁
Is required. Then, based on the deviation angle .delta..theta ₁
And as it can be obtained deviation size ΔX from distance and the imaging angle and the like between the CCD camera 16 and the coated surface 12, after all, the deviation size ΔX in this case, the deviation size [delta] X ₁ on the image P as a variable It is calculated according to the following equation (2). Note that the displacement dimension ΔY in the Y-axis direction is similarly calculated according to the following equation (3).

δX ₁ / P _X = δθ ₁ / θ (1) ΔX = f ₁ (δX ₁ ) (2) ΔY = f ₂ (δY ₁ ) (3) As shown, there is a displacement of dimension ΔZ only in the Z-axis direction,
When the D camera 16 is located in the XZ plane,
As shown in FIG. 8, the images T 'and M' on the image P are respectively δX ₁ , δX ₂ from the center O in the X-axis direction.
Just shift. These deviation dimensions ΔX ₁ and ΔX ₂
16 correspond to the deviation angles δθ ₁ and δθ ₂ from the center of the visual field, respectively. Further, since the distance 1 from the CCD camera 16 to the mark M of the illuminant 10 ° C. reflected on the painted surface 12 changes in accordance with the deviation dimension ΔZ, the size of the image M ′ on the image P is changed.
a _X ′, a _Y ′ also changes from the target dimensions a _X , a _Y. Therefore,
The displacement dimension ΔZ in this case is the displacement dimension δX ₁ on the image P,
[delta] X _2, is calculated in accordance with the following formula for the dimensional error .delta.a _X variable (4). Incidentally, it may be used .delta.a _Y instead of dimensional error .delta.a _X.

ΔZ = f ₃ (δX ₁ , δX ₂ , δa _X ) (4) Further, as shown in FIG. 9, the luminous body 10 and the luminous body 10 have an angular deviation of α _Y only around the Y axis. CCD
When the camera 16 is positioned in the XZ plane, the image M 'on the image P is, as shown in FIG.
The center O is shifted by δX _{2 in the} X-axis direction. This deviation dimension δ
X ₂ corresponds to a deviation angle Δθ ₂ from the center of the field of view of the CCD camera 16. Further, the posture of the illuminant 10 ′ reflected on the painted surface 12 with respect to the CCD camera 16 and the distance between the mark M and the CCD camera 16 change, and the dimensions a _X ′,
a _Y ′ changes from the target dimensions a _X and a _Y corresponding to the shift angle α _Y. Accordingly, the deviation angle alpha _Y in this case, the deviation size [delta] X ₂ on the image P, is calculated according to the following equation to dimensional errors .delta.a _X variables (5). Although it is also possible to use a .delta.a _Y instead of dimensional error .delta.a _X, since the direction of dimensional error .delta.a _X from the orientation of the light emitter 10 'increases, it is desirable to use .delta.a _X. Regarding the shift angle α _X about the X axis,
Similarly, it is calculated according to the following equation (6).

α _Y = f ₄ (δX ₂ , δa _X ) (5) α _X = f ₅ (δY ₂ , δa _Y ) (6) FIG. 11 shows the positional deviation and angular deviation of the robot hand 22. And the positional deviation of the images T 'and M' on the image P and the dimensional error of the image M '. Also,
The above equations (2) to (6) are for the case where the positional deviations ΔX, ΔY, ΔZ and the angular deviations α _X , α _Y occur independently, but in practice, these positional deviations and angular deviations Occur in a composite manner, the shift dimensions ΔX, ΔY, ΔZ and the shift angles α _X , α _Y are comprehensively calculated in consideration of the other shift dimensions and shift angles. For example, displacement δ on image P
As is apparent from FIG. 11, X ₁ is generated in common with the displacement ΔX in the X-axis direction and the displacement ΔZ in the Z-axis direction. Therefore, when calculating the displacement dimension ΔX, the displacement dimension δX ₁
In addition, it is necessary to consider not only the displacement dimension ΔZ. Furthermore, since the deviation dimension δX ₁ is strictly affected by the angle deviation α _Y , when the deviation dimension ΔX is determined with higher accuracy, the deviation angle α _Y must be considered. .

In the series of signal processing performed by the image processing device 36, the part that executes step S7 corresponds to a shift amount determining unit.

Then, such steps S1 to S7 are repeated N times, and when the deviation dimensions ΔX, ΔY, ΔZ and the deviation angles α _X , α _Y are obtained N each, in step S9, the average value and standard value thereof are obtained. Statistical processing for obtaining a deviation or the like is performed, and teaching position correction data in the X-axis direction, the Y-axis direction, the Z-axis direction, around the Y-axis, and around the X-axis is created. After that, the operator confirms the teaching position correction data with the personal computer 32,
When it is determined that the teaching is to be corrected in accordance with the teaching position correction data, the teaching position correction data is output to the robot controller 34 in step S10, and the correction data is converted into the amount of displacement of each axis of the robot 30. To correct the position and posture of the robot hand 22
The teaching position data stored in 32 is rewritten.

As described above, in the coating surface inspection apparatus according to the present embodiment, the deviation dimension δX ₁ of the images T ′ and M ′ of the teaching point T and the mark M on the image P captured by the CCD camera 16 from the image center O. , δY ₁ , δX ₂ , δY ₂ and the dimensional errors δa _X , δa _Y of the image M ′, the deviation dimensions ΔX, ΔY, ΔZ from the reference position of the robot hand 22 and the deviation angles α _X , α _Y The position deviation of the robot hand 22 is corrected in accordance with this, so that the correction is automatically performed compared to the conventional case where the worker corrected the position deviation and the angle deviation by trial and error. It can be done in a short time and with high accuracy.

As mentioned above, although one Example of this invention was described in detail based on drawing, this invention can be implemented in another aspect.

For example, in the above-described embodiment, the case where the present invention is applied to the teaching correction of the device for inspecting the smoothness of the painted surface 12 and the like has been described. However, other robots such as a sealing robot and a welding robot are also suitable for teaching locations. The present invention can be applied at the time of teaching correction by providing a suitable reflecting surface.

Further, according to the present invention, since the displacement dimension and the displacement angle are quantitatively obtained, the positioning accuracy of the robot is evaluated based on the displacement size and the displacement angle and the displacement of the second reference point in synchronization with the movement of the robot. The present invention can be used for other than teaching correction, such as detecting the shift of the light emission timing.

Further, in the above embodiment, the deviation dimension Δ of the robot hand 22
Although X, ΔY, ΔZ and the deviation angles α _X , α _Y are calculated by an arithmetic expression, they may be read from a data map created in advance.

In the above-described embodiment, the dimensional errors δa _X and δa _Y of the image M ′ of the mark M provided on the striped plate 26 are detected. ,
The dimensional error of the image T 'of the teaching point T may be detected.

Further, in the above embodiment, the mark M as a second reference point is provided on the striped plate 26 of the luminous body 10 that emits scattered light. What is necessary is just to be provided so that it may be reflected.

In the above-described embodiment, whether or not the teaching is to be corrected is determined by the operator. However, the teaching may be automatically corrected according to the correction data.

Further, in the above embodiment, the image processing in the case where the light emitter 10 and the CCD camera 16 are positioned in the XZ plane has been described, but the robot hand 22 is rotated by 90 ° about the Z axis and the light emitter 10 is rotated. Alternatively, the CCD camera 16 may be positioned in the YZ plane, and the position shift and the angle shift of the robot hand 22 may be detected based on two types of images captured at those two positions.

Further, in the above embodiment, the shift dimensions ΔX, ΔY, ΔZ and the shift angles α _X , α _Y are detected, but it is not always necessary to detect all of them.
Depending on the type of robot no problem just detecting the part of it such as such only [Delta] X, [Delta] Y and alpha _X.

Although not specifically exemplified, the present invention can be implemented in various modified and improved modes based on the knowledge of those skilled in the art.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a painted surface inspection apparatus to which the present invention is applied. FIG. 2 is a view for explaining an example of an image picked up by a CCD camera in the apparatus shown in FIG. FIG. 3 is a view showing a striped plate provided on the luminous body of the apparatus of FIG. FIG. 4 is a flowchart for explaining the operation for correcting teaching in the apparatus shown in FIG. FIG. 5 is a view for explaining a displacement in the X-axis direction in the apparatus shown in FIG. FIG. 6 is a view for explaining an image captured by the CCD camera in the case of FIG. FIG. 7 is a view for explaining a positional shift in the Z-axis direction in the apparatus of FIG. FIG. 8 is a view for explaining an image picked up by the CCD camera in the case of FIG. FIG. 9 is a view for explaining an angle shift around the Y axis in the apparatus of FIG. FIG. 10 is a view for explaining an image picked up by a CCD camera in the case of FIG. FIG. 11 is a diagram showing a relationship between the positional deviation and the angular deviation of the robot hand in the apparatus of FIG. 1, and the positional deviation and the dimensional error of the teaching point and the mark on the image. FIG. 12 is a schematic configuration diagram of a conventional painted surface inspection apparatus. FIG. 13 is a diagram showing teaching points provided on a painted surface. FIG. 14 is a view showing a probe used when teaching the apparatus of FIG. FIG. 15 is a view showing a striped plate provided on the luminous body of the device of FIG.
FIG. 16 is a flow chart for explaining the operation when correcting teaching in the apparatus shown in FIG. 12: Painted surface (reference surface) 16: CCD camera (imaging device) 22: Robot hand (movable part) 26: Striped plate (second reference member) 30: Robot, 36: Image processing device T: Teaching point (No. M: mark (second reference point) P: image O: image center (first target position, second target position) T ': first reference point on image M': second reference point on image Point Step S2, S3: First displacement detection means Step S4, S5: Second displacement detection means Step S6: Dimension error detection means Step S7: Displacement amount determination means

Claims (1)

(57) [Claims] A movable part of a robot which is relatively moved with respect to a mirror-like reference plane reflecting light includes a posture with respect to a first reference point provided on the reference plane, with respect to the reference plane. A second reference point is provided at a predetermined position, wherein the second reference point is provided at a predetermined position.
A second reference member disposed at a fixed position on the movable portion so that the reference point is reflected on the reference surface; and the movable portion is moved to a position near the reference position, thereby being reflected on the reference surface. An imaging device fixedly disposed on the movable unit so that the second reference point enters the imaging field of view together with the first reference point; and the first imaging device on the image captured by the imaging device.
A first step of detecting a deviation dimension between a reference point and a predetermined first target position where the first reference point is located on the image when the movable portion is located at the reference position;
Position deviation detecting means, the second reference point on the image, and a second predetermined reference position where the second reference point is located on the image when the movable portion is located at the reference position. A second positional deviation detecting means for detecting a deviation size from a target position, a size of at least one of the first reference point and the second reference point on the image, and the movable portion being located at the reference position. Dimensional error detecting means for detecting an error from a predetermined target size as the size of the one reference point on the image when the first position shift is detected, the first position shift detecting means, the second position shift detecting means, A displacement dimension and a displacement angle of the movable portion with respect to the reference position are determined based on the displacement error and the displacement error on the image detected by the displacement error detecting means. Positional deviation detecting apparatus for a robot, characterized in that it comprises a Ruzure amount determining means.