JPH054605B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention combines a light projector that projects slit light onto an object such as a welded joint, and an imaging device that takes an image of the slit light developed on the object as a sensor head. The present invention relates to an automatic work line detection method for detecting work lines such as welding lines by a light cutting method using a visual sensor provided as an optical sensor.

[Conventional technology]

FIG. 6 is a diagram showing an example of the conventional basic system configuration of this type of device, in which 1a and 1b are welding base materials to be overlap welded, and 2 is a welding line (work line). 3 is a slit light projector, and 4 is an imaging device, both of which constitute a sensor head 5 of a visual sensor.The image signal (video signal) sent by the imaging device 4 is sent to a signal processing device 6 including a microprocessor. Supplied. The slit light projector 3 projects slit light 3a in a direction crossing the welding line 2 to create a slit light image 3b on the welding base material.
The imaging device 4 captures the slit light image 3b,
The image signal is supplied to the signal processing device 6. The signal processing device 6 is, for example, a captured image (shown in FIG. 7) transmitted from the imaging device 4 at a predetermined detection cycle (sensor detection cycle). The image data of image 3c) F is captured, and from the captured image data, the brightest point in the captured image F is detected repeatedly in the x direction as the light cutting line position, and the brightest point detection data (slit light image 3c) is obtained. position data) is created, and from this position data, the position of the bending point B is detected as the position of the welding line 2, and sent to a torch drive device (not shown).

[Problem that the invention seeks to solve]

In this way, in conventional work line detection using a visual sensor, for example, the brightest point in the captured image F is detected as a light cutting line, and the work line is detected based on this light cutting detection position data. Since the cutting line position detection process is performed for the entire area of the captured image F, that is, for all of the captured image data, it takes less time to detect the work line after capturing the image data of the captured image F. In the case of automatic welding equipment, this has not only been a cause of impeding welding speed, but also disturbance light such as spatter light enters the field of view of the imaging device, causing a spatter trajectory S (Fig. 8) to appear in the photographed image F. In this case, erroneous detection is likely to occur at the intersection of the optical cutting line 3c and the sputter locus S, which is a factor that hinders the improvement of reliability.

This invention was made in order to solve the above-mentioned conventional problems, and aims to provide a work line detection method that can greatly increase the processing speed of automatic processing equipment and improve reliability. purpose.

[Means to solve the problem]

In order to achieve the above object, the present invention creates a window that limits the work line detection processing area for the currently captured image, on the condition that there is a work line position detected and stored in the previously captured image. In the second invention, the sensor head is further made rotatable around the axis of the processing tool of the automatic processing device in which it is mounted, and the work line position detected and stored this time is set in a preset position. If it deviates from the center line of the photographed image by exceeding the threshold, rotate the sensor head to eliminate the deviation, and change the currently detected and memorized work line position to the work line on the photographed image at the rotating position. The configuration is such that it can be rewritten to the position.

[Effect]

In this invention, since the work line detection process is performed only on the image data within the window, which is a small area compared to the entire photographed image, the light cutting line detection position data (position data of the slit light image) is created. This reduces the time required to detect the working line position from the brightest point detection data, which greatly improves the processing speed of automatic processing equipment compared to conventional methods. Ru. In addition, by setting the window, it is possible to exclude the intersection of the optical cutting line and the spatter locus from the optical cutting line position detection processing, which reduces the influence of ambient light on detection accuracy, increasing reliability. can be increased. Additionally, if the work line deviates from the center line of the photographed image by more than a position threshold, the sensor head rotates to eliminate this deviation, so the above effect can be obtained even when the work line is a curved line. .

〔Example〕

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

In FIG. 1a, 10 is a welding torch, 11 is a wire, and 12 is a sensor head of a visual sensor, which has a slit light projector 3 and an imaging device 4. As shown in FIG. 1b, a welding torch is attached to a torch holder 13. It is supported so as to be able to rotate around the 10 torch shafts T as rotation centers. 14 is a signal processing device that processes an image signal from the imaging device 4 to detect the welding line position; 15 is a sensor rotation mechanism for rotating the sensor head 12; 16 is the sensor rotation mechanism 1
5, which receives a drive signal from the signal processing device 14 and rotates the sensor rotation mechanism 1.
Drive 5.

Next, the operation of this device will be explained with reference to the processing flow shown in FIG.

When detecting the starting point, the signal processing device 14 sets the entire captured image corresponding to the entire field of view of the imaging device 4 as a "welding line detection processing area" and starts welding line detection processing. This is the starting point of welding line 2 (A in Figure 7).
This is because when detecting point A, it is not known at what position within the field of view the field of view of the imaging device 4 will capture the welding start point A due to setting errors in the welding base material. The signal processing device 14 processes the captured photographic image.
When the welding line detection process (described later) is executed on the image data of F _i to detect the starting point of welding line 2 (welding starting point), the position of this welding starting point is stored, and a robot, etc. that controls the position of the welding torch 10 as a control signal. As a result, the welding torch 10 is controlled to the starting point A of the welding line, and welding work is started.

Signal processing device 1 that detected and memorized the above welding start point
4 subsequently captures a new captured image F _i , but for the captured image F _i (i=1, 2, 3...), the third
As shown in the figure, a "welding line detection limited processing area" (hereinafter referred to as window) W is set. This window W is, for example, a small area that is similar to the field frame of the imaging device 4, and has a base material shape,
The moving speed of robots, etc., the detection cycle of visual sensors,
It is determined by taking into account thermal deformation distortion of the base material in the detection period, and a window W is created at or near the position of the welding line 2 detected and stored in the previously captured photographic image F _i-1 as a reference. is set so that the center of This situation is shown in Figure 3 (a and b).
It is shown in two diagrams. Welding line 2 has continuity,
Since the change in the position of the welding line in the photographed image for each sensor detection cycle is minute, the window W is set at a position that reliably surrounds the welding line 2 in the photographed image F _i . The signal processing device 14 uses this window W.
After setting, welding line detection processing is executed within this window W. That is, for the image data within the window W, the light section line position is sequentially and repeatedly detected in the x direction as a bright part having a light section line extraction brightness threshold or higher set in advance, and the slit light image 3b is detected within the window W. Position data of the image portion is created, and from this position data, the position of the above-mentioned refraction point B is detected as the position of the weld line 2 and stored. From then on, the signal processing device 14 uses the captured image F _i (i=n) until it detects the end point of the welding line 2.
, the previously captured photographic image F _i (i=n−
A window W is set near the position of the welding line 2 detected in step 1), and welding line detection processing is performed on the image data within this window W.

In this way, in this embodiment, each time a captured image is captured, the window W is set and the welding line detection processing area is limited to a small area, so the processing time required for the optical cutting line detection process can be shortened. , as the optical cutting line detection position detection data also decreases,
Since the weld line detection processing time for processing optical cutting line detection position detection data to determine the weld line is also shortened, the overall processing time is significantly shorter than before, making it possible to significantly increase welding speed. do. In addition, by setting the window W, the intersection between the disturbance light image S and the slit light image 3c can be excluded from the light cutting line position detection process, so the probability of being affected by the disturbance light is reduced. This prevents deterioration in detection accuracy due to ambient light.

Further, the signal processing device 14 monitors the position of each welding line for each photographed image, and when the position of the welding line 2 exceeds the position threshold, the image center line M of the photographed image F (the center line of the field of view of the imaging device 4 ), the sensor head 12 is rotated in a direction to eliminate the deviation so that the weld line 2 is located on or near the center line M. That is, as shown in FIG. 4, the angle _θ i made with the visual field center line M at the two positions of the weld line detected in the captured image F _i
is the preset position threshold indicated by the imaginary line |
If it exceeds θ _TH |, a turning command (signal) for rotating the sensor head 12 by an angle θ _i in the direction of the welding line 2 is generated and sent to the drive device 16. As a result, the sensor head 12 is rotated, so that the weld line 2 is aligned with the center line M as shown in FIG.
It will be located above or nearby.

Therefore, the welding progresses to the curved part of the welding line, and as the welding progresses, the welding line 2 within the field of view of the imaging device 4 gradually shifts to one side within the field of view. Even in a case where the welding line 2 deviates from the window W when the welding line 2 is fixed, the welding line 2 can be reliably captured within the window W according to this embodiment.

In this way, when the sensor head 12 is rotated, the position of the field of view of the imaging device 4 with respect to the base material changes, so that the position of the weld line 2 in the photographed image F _i (shown by dotted lines in Figure 5) before the head is rotated is the same as that of the head. The position corresponding to the photographed image F _i+1 after turning is determined based on the weld line position and the turning angle θ _i , and the memorized position (welding line 2 in the photographed image F _i before the head turns)
) to the newly found position above,
The window W of the captured image F _i+1 is then set based on this rewritten position. Although the above-mentioned embodiment describes the case of detecting a weld line, the present invention can be applied to detecting a work line of other automatic processing devices such as an automatic sealing device to obtain the same effect.

〔Effect of the invention〕

As explained above, this invention sets a window that limits the work line detection processing area to a small area for each photographed image during copy processing by an automatic processing device, so the work line detection process can be performed faster than in the past. In addition to being able to perform position detection, the detection accuracy can be improved.Furthermore, when the work line position exceeds the position threshold, the sensor head is rotated to correct the field of view. Not only when is a straight line, but also when it includes a curve or a curved part,
The above effects can be obtained.

[Brief explanation of drawings]

FIG. 1a is a block configuration diagram showing an embodiment of the present invention, FIG. 1b is a diagram showing the support structure of the sensor head in the above embodiment, and FIG. 2 is a flowchart showing the weld line position detection processing flow in the above embodiment. figure,
FIG. 3 is a diagram for explaining the window in the above embodiment, FIGS. 4 and 5 are diagrams showing photographed images for explaining the correct response action of the sensor head in the above embodiment, and FIG. 6 is a diagram for explaining the conventional working line. Block diagrams for explaining the position detection method, FIGS. 7 and 8
The figure is a diagram showing an example of a photographed image. 3... Slit floodlight, 4... Imaging device, 10
...Welding torch, 12...Sensor head, 14...
. . . signal processing device, 15 . . . sensor head rotation mechanism, 16 . . . drive device for the sensor head rotation mechanism.

Claims (1)

[Claims] 1. Using a sensor head that has a projector that projects slit light onto an object and an imaging device that takes an image of the slit light developed on the object, images are taken at a predetermined period from the imaging device. In the automatic work line detection method for detecting work lines of automatic processing equipment equipped with the above-mentioned sensor head, a window that limits the work line detection processing area to a small area is created in the previously captured image for the currently captured image. A work line detection method characterized in that, on the condition that there is a detected and stored work line position, a position surrounding the work line position is set, and a work line detection process is executed for data in the window. 2 Using a sensor head that has a light projector that emits slit light onto the target object and an imaging device that captures the slit light image developed on the target object, images are captured from the imaging device at a predetermined period, and the sensor head is mounted. In the automatic working line detection method for detecting the working line of an automatic processing device, the sensor head is made rotatable around the processing tool axis of the automatic processing device, and the working line detection processing area is set for the captured image. A window that is limited to a small area is set to a position that surrounds the work line position that was detected and stored in the previously captured captured image, and is also set to a position that surrounds the work line position that was detected and stored this time. If it deviates from the center line of the photographed image by exceeding a preset position threshold, the sensor head is rotated to eliminate the deviation and the currently detected and memorized working line position is changed to the photographed image at the swivel position. An automatic work line detection method characterized by rewriting the work line position to the upper work line position.