JP2661809B2 - Position detection method for automation - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a position detecting method for automatically detecting the position of a work object, which is a position object, in an environment such as high temperature and dust.

[0002]

2. Description of the Related Art In general, when a specific operation is performed by an industrial robot, the position of an object to be processed is detected using a camera device. Conventionally,
As a method of detecting the position of such a work target,
With a camera device using a solid-state imaging device (CCD),
In addition to photographing the work object, the image is image-processed, for example, binarized, and the position of the work object is detected.

[0003]

In the case of automation or unmanned operation in an environment such as high temperature and dust in a steel mill or the like, a camera device and an image processing device as described above are used. When detecting the position of an object, there are the following problems.

[0004] The environment inside the steel works is dim as a whole, and the work object and its surroundings have the same color due to the adhesion of soot by fire, the accumulation of dust, the generation of rust, and the like. There is little contrast between light and dark between the work object and the surrounding image, and there is a case where a clear shape of the work object cannot be obtained even by image processing, and therefore, the position of the work object cannot be accurately detected. There was a problem.

[0005] In steel works, one operation is performed at intervals of 30 minutes to one hour, the frequency of repetition is low, and one robot is often in charge of several types of operations. Therefore, since the shape, color, and lighting state of the work object are different for each work, it is necessary to find an image processing method that enables position detection for each work object, which requires a great deal of time and effort. There was a problem.

Accordingly, an object of the present invention is to provide a position detection method for automation which can solve the above problem.

[0007]

In order to solve the above problems, a first means of the present invention comprises a plurality of ceramics provided at a predetermined distance from a position detection target.
The obtained detection member is photographed by the camera device, the image obtained by the camera device is subjected to image processing by the image processing device, and the position of the position detection target is determined based on the detection member position obtained by the image processing. When detecting,
This is a position detection method for automation in which the color of the detection member has contrast with the surrounding color.

[0008]

Further, in order to solve the above-mentioned problem, a second means of the present invention provides a method in which a color having a contrast with respect to the color is applied to the periphery of the position detection object and the
The detecting member constituted by click, arranged such that the position detection object to the center position of its comes, the image while photographing the detected member by a camera device, the image obtained by the camera device by the image processing apparatus This is a position detection method for automation in which the position of the position detection target is detected by performing processing and determining the center position of the detection member.

[0010]

According to the above-described position detection method, a detection member having a contrast with respect to the surroundings is arranged at a position different from the position detection target, and the detection member is photographed and the image is processed. Since the center position of the position detection target is determined, the position of the position detection target can be detected very accurately even in a bad environment.

Further, by cleaning the surface of the detection member with air, accurate position detection can be performed even in a poor environment where soot and dust are generated. Further, a method of arranging a detection member having a contrast around the position detection target and detecting the detection member by the same image processing as described above is also performed. In this case, the position of the position detection target is also accurately detected. be able to.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described below with reference to FIGS. In FIG. 1, reference numeral 1 denotes an industrial robot (hereinafter simply referred to as a robot) for automatically performing various operations in a steel mill, for example, installing a long nozzle attached to a ladle or a supply hose such as air or hydraulic pressure. In order to perform such work automatically, when grasping the tip of the long nozzle or the coupling attached to the tip of the supply hose with the gripping part of the robot, the position of each tip is Need to be detected.

In this embodiment, a description will be given of a method of detecting the position of the opening at the tip of a long nozzle when the oxygen cleaning operation is automatically performed by a robot. The robot 1 controls the support 2, the arm 3 supported by the support 2, the grip 4 attached to the tip of the arm 3, and the arm 3, the grip 4, and the like. And a control device 5 that performs the control.

A camera device 6 using a solid-state image pickup device (CCD) is attached to the tip of the arm portion 3 of the robot 1, and an image taken by the camera device 6 is input and a predetermined image is input. An image processing device 7 for performing image processing is provided.

Numeral 11 is a support member for the long nozzle 12, and its horizontal support portion (for example, the tip is forked) 1
1a, the long nozzle 12 is supported via a plate-shaped receiving member 13 attached to the tip of the long nozzle 12.

The receiving member 13 attached to the tip of the long nozzle 12 has a pair of right and left at a position separated from the tip opening 12a of the long nozzle itself by a predetermined distance (the accuracy is maintained by machining). (Detection members) 14A and 14B are disposed such that their surfaces are exposed and embedded (note that in FIG. 1, the surfaces are projected so as to be clear for clarity). ). Of course, it may be arranged so as to protrude from the surface of the receiver 13.

The pins 14A and 14B themselves are made of a white heat-resistant ceramic, and the surrounding receiver 13 is made black, so that the contrast of the pins 14A and 14B with the surroundings becomes remarkable. Is taken into account.

The center position of each of the pins 14A and 14B and the center position of the tip opening 12a of the long nozzle 12 are accurately determined in advance. That is, the center of the distal end opening 12a of the long nozzle 12 is located on a vertical bisector B of a straight line A connecting the centers of both pins 14A and 14B and at a position separated by a predetermined distance C from the straight line A. It is set to go.

Further, each pin 14 on the receiving member 13
At positions corresponding to A and 14B, an air injection device (consisting of a nozzle portion 16, a compressed air source (not shown), etc.) for blowing off soot and dust attached to the pins 14A and 14B with cleaning air. 15 is provided via a support base (not shown).

Note that the air injection device 15 is
3, or may be attached to the arm 3 of the robot 1. Next, the position detection procedure will be described.

First, the long nozzle 12 to which the receiving member 13 is attached is supported by the horizontal support portion 12a of the support member 11, and then each of the pins 14A, 1
The cleaning air is sprayed on the surface of 4B to remove soot and dust.

Next, two pins 1 are
4A and 14B are taken, and the images are input to the image processing device 7 and subjected to image processing such as binarization processing to obtain pins 14A and
The image of only 14B is extracted.

Next, the center coordinates of the extracted images of the pins 14A and 14B are obtained by image processing.
The center position of the distal end opening 12a of the long nozzle 12 is obtained based on the coordinates of 4A and 14B.

That is, after a vertical bisector B of the line A connecting the centers of the two pins 14A and 14B is determined, the vertical bisector B is separated from the line A by a predetermined distance C. Point D
Is the position of the end opening 12a of the long nozzle 12 to be obtained.

As described above, instead of directly detecting the position of the work object, the position of the work object is detected by detecting the position of two pins provided for position detection. Has the following effects.

When the illuminance in the working environment and the shape and color of the work object are different, there is no need to create image processing software in accordance with the illuminance and the work object.

Even if an erroneous detection occurs, it can be dealt with only by changing the size and color of the position detecting pin. In addition, since the position detection pin is made of white ceramic and its surface is cleaned with air, when the contrast between the work object and its surroundings is low, and where there is a lot of soot and dust, If it is below, or if you are in a hot environment,
An image having a clear contrast was obtained, and thus the erroneous detection rate of the position of the work object could be reduced.

As described above, the long nozzle 12
When the position of the tip opening 12a of the long nozzle 12 is detected, the position data is output to the control device 5 of the robot 1, and the oxygen cleaning nozzle (not shown) is gripped by the robot 1, and the oxygen cleaning of the long nozzle 12 is performed. Work is performed.

Next, a second embodiment of the present invention will be described with reference to the drawings. In the first embodiment, the position of the tip opening of the long nozzle is detected. In the second embodiment, for example, hydraulic pressure (or air pressure) is applied to a sliding nozzle provided at the bottom of the ladle. A method for detecting the position of the quick coupling attached to the tip of the supply hose to automatically attach and detach the supply hose will be described.

That is, as shown in FIGS. 4 and 5,
A quick coupling (in this case, male, but of course, female) may be provided at the tip of the hydraulic supply hose 21.
2 is attached, and a circular positioning detection plate (detection member) 23 made of a white heat-resistant ceramic is attached to a base portion thereof.

Of course, the detection plate 23 and the quick coupling 22 are arranged such that the center position of the detection plate 23 coincides with the center position of the quick coupling 22. Next, a method for detecting the position of the quick coupling will be described.

When light is irradiated from one direction in a dark environment, as shown in FIG. 6, half of the quick coupling 22 itself appears dark. In FIG. 6, dark portions are indicated by oblique lines.

When this state is photographed by the camera device as described in the first embodiment, the image cannot completely recognize the shape of the quick coupling 22, as shown in FIG. Although the image appears white and the other half appears black as indicated by the diagonal lines), the periphery of the detection plate 23 is clearly recognized as white.

Therefore, image processing is performed by the image processing apparatus, and as shown in FIG. 7, the black portion surrounded by white is filled with white, and the center coordinates of the white circle are formed by image processing such as face center calculation. Ask for.

When only the quick coupling is photographed and binarized as in the prior art, as shown in FIG. 8, a black portion of the shape of the quick coupling 22 is missing as shown in FIG. Therefore, the accurate position of the quick coupling cannot be detected.

Thus, by arranging the work object at the center of the detecting member having a remarkable contrast with respect to the surroundings and detecting the detecting plate which is easy to detect,
The position of the work object can be accurately detected.

In each of the above embodiments, the detection members such as the pins and the detection plates are described as being white.
The color is not limited to white, and may be any color that makes the image clear when the detection member is photographed by a camera device, that is, a color that provides the best contrast to the surroundings.

The shape of the pin or the detection plate is not limited to a round shape, but may be, for example, a rectangular shape.

[0039]

As described above, according to the position detection method of the present invention, a detection member having a contrast with respect to the surroundings is arranged at a position different from the position detection target object, and the detection member is photographed. By processing the image,
Since the center position of the position detection target is determined, the position of the position detection target can be detected very accurately even in a bad environment, and the software for image processing can also be used for the position detection target. It is not necessary to change for each position detection target.

Further, even when a detection member having a contrast is disposed around the position detection target, accurate position detection of the position detection target can be performed as described above.

[Brief description of the drawings]

FIG. 1 is an overall schematic side view showing a first embodiment of a position detecting method according to the present invention.

FIG. 2 is a main part plan view for explaining a position detection method in the first embodiment.

FIG. 3 is a sectional view of an essential part for explaining a position detecting method in the first embodiment.

FIG. 4 is a main part side view showing a second embodiment of the position detection method of the present invention.

FIG. 5 is a plan view of a main part showing the second embodiment.

FIG. 6 is a plan view illustrating a position detection method according to the second embodiment.

FIG. 7 is a main part plan view for explaining a position detection method in the second embodiment.

FIG. 8 is a plan view of a main part for explaining a conventional position detection method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Industrial robot 3 Arm part 4 Grasping part 5 Control device 6 Camera device 7 Image processing device 11 Support member 12 Long nozzle 13 Receiver 14A, 14B Pin 21 Hydraulic supply hose 22 Quick coupling 23 Detection plate

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification code Agency reference number FI Technical display location G06T 7/60 G06F 15/70 350B (72) Inventor Wataru Noguchi 11-1 Showa-cho, Kure City, Hiroshima Prefecture No. Nisshin Steel Corporation Kure Works (72) Inventor Takashi Tsubo 11-1 Showa-cho, Kure City, Hiroshima Prefecture Inside Nisshin Steel Corporation Kure Works (72) Inventor Kazuhiro Sawamura Showa-cho, Kure City, Hiroshima Prefecture No. 11 Inside Nisshin Steel Corporation Kure Works (56) References JP-A-1-174902 (JP, A) JP-A-60-242305 (JP, A) JP-A-59-154305 (JP, A) )

Claims (2)

(57) [Claims] 1. A camera device is used to photograph a detection member made of a plurality of ceramics provided at a predetermined distance from a position detection target object, and an image obtained by the camera device is image-processed by an image processing device. Processing, when detecting the position of the position detection target object based on the detection member position obtained by this image processing, the color of the detection member is given a contrast with the surrounding color. Position detection method for automation. Around 2. A position detection object, color having a contrast against the color is applied and the ceramic
More configured detecting member, arranged such that the position detection object to the center position of its comes, while imaging the detection member by the camera device, image processing an image obtained by the camera device by the image processing apparatus And detecting the center position of the detection member to thereby detect the position of the position detection target object.