JPH085350A - Weld groove position and shape measurement method - Google Patents

Abstract

PURPOSE:To stably and accurately measure a groove position and a groove shape even if the groove is provided with a surface state which is subjected to grinding finish by synthesizing emission line position data from left and right directions of a slit image. CONSTITUTION:A slit light projector 8 and a CCD camera 9 are installed at a welding head 7 at a certain angle, thus picking up the image of the reflection light of slit light 13 which is applied to a groove 12. An obtained slit image 14 is taken into an image processing device. Then, an emission line position is determined by tracking a right direction from the left edge of the image 14 and the emission line position is determined by tracking a left direction from the right edge of the image 14. The emission line position data from the left and right directions are synthesized and proper emission line position data are determined. The emission line position data are changed to curvature data by operation and a feature point is extracted from emission line curvature data and the brightness distribution data at the bottom of the groove 12. Then, the position and shape (groove width, route gap, dislocation between plates etc.) of the groove 12 can be calculated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to automatic welding of steel materials, and more particularly to a method for rapidly and automatically measuring the shape of a so-called groove portion where welding is performed.

[0002]

2. Description of the Related Art When different steel materials are butted and welded together, the butted portion usually has a V, K or I shape and is called a groove (hereinafter referred to as a groove). V
Explaining with an example of a groove of a shape, as shown in FIGS. 15 (a) and 15 (b), it is from a portion called groove width 1, groove depth 2, groove angle 3, misalignment 4, 4 '. In addition to this, there is a so-called root face 5 joining surface between steel materials and a joint surface gap called a root gap 6. Further, it is important for the success of automatic welding to know the groove position and the shape thereof promptly and surely, which is not widely used in recent years. So, conventionally,
A method and apparatus for automatically measuring the position and shape of the groove have been developed and studied, and some of them have been put into practical use.

For example, as an early technique, Japanese Patent Laid-Open No. 52-
JP-A-133050 discloses an optical device for irradiating a groove with slit light, an image pickup device for picking up a slit image showing a groove cross-sectional shape by the slit light, a storage device for storing the taken slit image, and It is composed of a coordinate extraction device that replaces necessary data among the stored slit image data into a two-dimensional XY coordinate system using a computer, and an estimation device that estimates the groove position and shape from the coordinate values. It is a thing. However, in the slit image measured by the device, there is a problem that a groove inclined portion is erroneously detected by an unnecessary reflection image due to multiple reflection of slit light on the groove surface or an arc light transmitted through the groove surface. In the latter half of the multi-layer welding, the slit image data becomes very small, so that the accuracy of the linear approximation decreases and the accuracy of detecting the groove center position decreases.

To solve these problems, Japanese Patent Application No. 56-118604 discloses a slit image brightest point detection circuit in the above apparatus and smoothes and differentiates the image data detected by this circuit. Circuit and added. That is, the slit image is tracked by the brightest point detection circuit to provide a limited range in which unnecessary light does not exist, and the groove center position is detected only by the slit images at the horizontal part of the groove and the upper end of the hypotenuse part.

However, when the above measuring device is applied to a groove having a surface state that has been grinder-finished in order to prevent welding defects, the photographed image has the following state due to the unique surface state of the groove. Become. 1) On the groove surface having an inclination angle with respect to the slit light irradiation direction, the blur of the slit bright line image is likely to occur. The slit bright line is formed by a plurality of spots of slit light (hereinafter, referred to as pixels). 2) The brightness of the slit bright line varies from area to area. 3) A multiple reflection image is reflected on the slope of the groove. In particular, this tendency becomes remarkable when the root gap of the groove is narrowed or the groove angle is narrowed to narrow the groove, and the left and right groove surfaces approach each other.

Therefore, the above-mentioned JP-A-56-118604 is used.
The device disclosed in the publication has the following problems in addition to being able to measure only the groove center position. a) It is difficult to determine the feature point at the lower end of the groove from the slit bright line image on the groove surface in the state of 1) and 3). Therefore, the root gap cannot be measured. b) When the information of the position of the slit bright line in the image (determined by the two-dimensional coordinates of each pixel) is determined by the tracking method, the multiple anti-mapping is erroneously tracked at the bottom of the groove, and correct slit bright line position data acquisition is performed. It often becomes impossible. However, if the tracking method is not used, all the pixels in the image are set as the processing target pixels, which causes a problem that the time required for the calculation becomes long, or the noise of the image becomes frequent. c) When the position of the slit bright line image is determined as the brightest point on the scanning line installed in the vertical direction of the image, the brightness distribution in the width direction of the slit bright line is non-uniform, and thus the determined bright line position data fluctuates finely. including. This fluctuation becomes noise when the groove feature points are extracted from the bright line position data. d) The smoothing differentiating method used as the data processing for extracting the characteristic points from the bright line position data is easily affected by the noise of the bright line position data generated by the adhesion of spatter or the flaws on the surface of the base material.

[0007]

SUMMARY OF THE INVENTION In view of the above situation, the present invention provides a method for opening even a groove having a grinder-finished surface state which makes it difficult to apply an image sensor combined with slit light irradiation. Tip position and groove shape (groove width, root gap, misalignment, groove angle, etc.)
It is an object of the present invention to provide a welding groove position and shape measuring method capable of stably measuring with high accuracy.

[0008]

In order to achieve the above object, the inventor reviewed the conventional slit image processing method, and earnestly studied the noise removal from the image. The present invention has been made as a result of the efforts, namely, irradiating the groove formed by the welding base metal with slit light, photographing the reflected light from the groove surface, and obtaining the slit image. In the method for measuring the position and shape of the welding groove to be image-processed, the slit image is sequentially processed in the steps described below, which is a method for measuring the position and shape of the welding groove. Step 1. Step 2. Capture of slit image in image processing device 2. Determine the bright line position by tracing from the left edge of the slit image to the right. 3. Determine the bright line position by tracing from the right end of the slit image to the left. Synthesize the bright line position data from the left and right directions,
Determination of appropriate bright line position data Step 5. Calculation of change from bright line position data to curvature data Step 6. Feature point extraction of groove bottom from bright line curvature data and luminance distribution data of groove bottom Step 7. Calculation of groove position and groove shape

[0009]

According to the present invention, the groove formed by the welding base material is irradiated with slit light, the reflected light from the groove surface is photographed, and the slit image thus obtained is subjected to image processing. In the measuring method, since the slit images are sequentially processed in the steps described below, it becomes possible to measure the welding groove position and shape with higher accuracy and accuracy than conventional ones. As a result, automatic welding can be performed smoothly. Step 1. Step 2. Importing the image into the slit image processing device 2. Determine the bright line position by tracing from the left edge of the slit image to the right. 3. Determine the bright line position by tracing from the right end of the slit image to the left. Synthesize the bright line position data from the left and right directions,
Determination of appropriate bright line position data Step 5. Calculation of change from bright line position data to curvature data Step 6. Feature point extraction of groove bottom from bright line curvature data and luminance distribution data of groove bottom Step 7. Calculation of groove position and groove shape The contents of the present invention will be specifically described below with reference to FIGS.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram of an apparatus for carrying out a welding groove position and shape measuring method according to the present invention.
It comprises a welding head 7, a slit light projector 8, a CCD camera 9 and an image processing device 10. The slit light projector 8 and the CCD camera 9 are installed so as to form a certain angle (here, the reflected light from the groove surface is inclined by 45 ° with respect to the slit light plane), and the groove 12 is irradiated with the light. The reflected light of the slit light 13 thus formed was imaged at this angle, and the position and sectional shape of the groove were measured from the obtained slit image 14 (hereinafter, simply referred to as image). The measuring method according to the present invention was performed by sequentially performing a plurality of steps shown in FIG. 2 using this apparatus.

First, as the first step, the slit image 14 is taken into the image processing apparatus 10. The slit image 14 is as shown in FIG. 3 (actual photographed image is shown in FIG.
6 is shown in the photograph). Then, in the second step, the position data of the bright line is acquired from the left side of the slit image 14 by the tracking method. The tracking method is a kind of extraction method of bright line data in image processing, and in this case, bright line position pixels (coordinates X _i , Y) that are adjacent in the X-axis direction in the X-Y coordinates.
_i ) are connected to create bright line position data (see FIG. 4). Here, i is a natural number of 1, 2, 3, ..., N, and n is x of the image.
It represents the number of constituent pixels in the direction. A specific method of continuing the tracking while determining the position of the bright line is as shown in the flowchart of FIG. 9 and FIGS. 10 and 11. According to this method, only the image data around the bright line is efficiently handled. You can As is apparent from FIGS. 9 and 11, in the present invention, the pixel (X _i, Y _i80U ) whose brightness is 80% of the _brightest point,
(X _i , Y _i80D ) is the edge (outline) of the bright line, and the midpoint thereof is the bright line position (X _i , Y _i ). Note that FIG.
The "threshold value" is a "binary process" in order to recognize the position on the image in the light-section method, that is, "white" when the brightness on the image exceeds a certain set value, and "white" otherwise. It is judged as "black", but this setting value. In FIG. 11, this value was set to 80% of the brightest point. However, in the tracking method of FIG. 9, if the image is bisected by the center line, scattered light may be tracked instead of the true slit bright line on the right groove surface, resulting in erroneous detection. In the third step, the position data of the bright line is acquired from the right side of the image according to the same principle as in the second step (see FIG. 5).

Next, in the fourth step, the bright line position data is combined with the left and right obtained in the second step and the third step. At that time, as is apparent from the dotted line in FIG. 6, in the area where the bright line data overlaps on the X axis, the bright line data having the larger Y value is determined as the true data. This corrects the erroneous detection of the tracking method of FIG. 9 described above, and is a judgment based on the condition that scattered light does not occur above the slit light irradiation unit.

Subsequently, in the fifth step, in order to facilitate the subsequent data processing, the bright line position data obtained in the fourth step is subjected to curvature calculation processing to create curvature data (see FIG. 7). The data before and after m pieces of pixel data (X _i , Y _i ) to be processed are (X _{i + m} , Y _{i + m} ) and (X _im ,
Y _im ), the curvature data θ at the processing target point
_As shown in FIG. 12, _i can be obtained by Equation 1 by the inner product method of the data of three points. Here, m is an integer set according to the groove shape.

[0014]

[Equation 1]

In the sixth step, as shown in FIG. 7, the X value, which is the maximum value of the curvature data θ _i , is determined as the existence of the left groove lower end feature point p ₁ and the right groove lower end feature point p ₂ . It is determined to be the groove bottom representative position p ₀ . Then, the accurate positions of p ₁ and p ₂ are determined based on the luminance distribution information (brightness / darkness information) around the p ₀ . The method of determining the positions of p ₁ and p ₂ is as shown in the flowchart and explanatory diagram of the groove bottom feature point position determining method shown in FIGS. 13 and 14. Also,
In the curvature data shown in FIG. 8A, the X value having the minimum point on the left side of p ₀ is determined as the X value at which the left groove shoulder feature point p ₃ exists, and the minimum point on the right side of p ₀ is determined. It is determined that the X value that takes is the X value at which the right groove shoulder feature point p ₄ exists.

Finally, FIG. 8 determined as described above
Each feature point p ₁ , p ₂ , p on the image as shown in (b)
_{From 3} and p ₄ , the groove position and the groove width 1 as the groove shape, the root gap 6, the misalignment 4, the groove angle 3 and the like are calculated. During welding, torch positioning is controlled according to these values and automatic welding is executed.

In the present embodiment, automatic MAG welding was carried out by abutting steel pipe joints having a wall thickness of 15.3 mm. At that time, both ends were ground into a chevron shape, and a so-called X-grinder polishing was performed. It was a bevel. The groove angle is about 40 °,
The root gap 6 was 0 to 0.5 mm. The slit light projector 8 irradiates the surface of the steel pipe perpendicularly, and the CCD
The camera 9 and the slit light plane were set at 45 °. As a result, erroneous detection of the groove position and shape due to reflected light or arc light as in the past was prevented, and the groove measurement during welding was almost 100% successful. Further, the detectability of the root gap 6 is 0.2 mm, the measurement accuracy of the groove position and shape is about 0.1 mm, and the time required for measurement for one image is about 0.2 seconds, which is very short. It is.

[0018]

As described above, according to the present invention, the groove position and groove shape of the welded portion can be measured with high accuracy and speed. As a result, the welding torch position can be accurately controlled by using these values, and automatic welding having excellent welds with few defects and good bead appearance has become possible.

[Brief description of drawings]

FIG. 1 is a perspective view showing a configuration of an apparatus for carrying out a welding groove position and shape measuring method according to the present invention.

FIG. 2 is a process drawing showing a method for measuring a welding groove position and a shape according to the present invention.

FIG. 3 is an example of a measurement image of a welding groove stored in an image processing device.

FIG. 4 is a diagram showing bright line position data obtained by scanning the measurement image of FIG. 3 from the left side.

5 is a diagram showing bright line position data obtained by scanning the measurement image of FIG. 3 from the right side.

FIG. 6 is a diagram showing appropriate data obtained by combining the bright line position data of FIGS. 4 and 5;

7 is a diagram showing data obtained by converting the combined data of FIG. 6 into curvature data.

8 is a diagram showing each characteristic point of a welding groove in the curvature data of FIG. 7, where (a) is curvature data and (b) is image data.

FIG. 9 is a flowchart illustrating a so-called tracking method in image processing.

FIG. 10 is a diagram showing scanning in a so-called tracking method.

FIG. 11 is a diagram illustrating a method of determining a bright line position in image processing, showing a luminance distribution of pixels existing in a vertical direction at a coordinate position X _i .

FIG. 12 is an explanatory diagram of an inner product method for converting bright line data obtained by image processing into curvature data.

FIG. 13 is a flowchart of a method for determining a feature point position of a groove bottom.

FIG. 14 is a diagram showing a result of carrying out the method of FIG.

FIG. 15 is a diagram showing characteristic points of each portion of a welding groove,
(B) shows a state in which the left and right sides of (a) are slightly shifted vertically.

16 is a photograph example of the slit image of FIG.

[Explanation of symbols]

 1 groove width 2 groove depth 3 groove angle 4 misalignment 5 root face 6 root gap 7 welding head 8 slit light projector 9 CCD camera 10 image processing device 11 welding target material 12 groove 13 slit light 14 slit image 15 Bright line 16 Scattered light 17 Scanning width

Claims (1)

[Claims] 1. A method for measuring the position and shape of a welding groove by irradiating a groove formed by a welding base material with slit light, photographing reflected light from the surface of the groove, and image-processing the obtained slit image. In the method of measuring the welding groove position and shape, the slit images are sequentially processed in the following steps. Step 1. Step 2. Capture of slit image in image processing device 2. Determine the bright line position by tracing from the left edge of the slit image to the right. 3. Determine the bright line position by tracing from the right end of the slit image to the left. Synthesize the bright line position data from the left and right directions,
Determination of appropriate bright line position data Step 5. Calculation of change from bright line position data to curvature data Step 6. Feature point extraction of groove bottom from bright line curvature data and luminance distribution data of groove bottom Step 7. Calculation of groove position and groove shape