JP3858792B2 - Method and apparatus for detecting bead shape of electric resistance welded pipe - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method and apparatus for detecting a weld bead shape of an electric resistance welded pipe. The bead shape referred to in the present invention includes the bead position.
[0002]
[Prior art]
In general, ERW welded pipe (hereinafter referred to as “ERW pipe” for short in the text), for example, ERW steel pipe is formed by forming a metal band (including a metal plate) such as a steel band into a tubular shape. While being conveyed in the longitudinal direction, both width ends of the metal strip such as the steel strip are continuously butt welded in the longitudinal direction by means such as high frequency induction heating pressure welding or resistance heating pressure welding.
[0003]
In the welded portion of the electric resistance welded pipe, a bulge by pressure welding, that is, a weld bead (hereinafter, also referred to as “bead” for short) occurs on the inner and outer surfaces of the pipe. Although this bead is cut in the course of pipe manufacture, it is conventionally known that the raised shape (width, height, etc.) of this bead before cutting is related to the weld strength at the final product stage. For this reason, conventionally, in the welding process, the worker adjusts the welding current and the like while visually observing the bead after welding. Since it is left to the subjectivity, there is a problem that differences between workers and temporal variations occur, and universality and reproducibility deteriorate. For this reason, attempts have been made to automatically detect beads by various methods. As for the invention related to the conventional method or apparatus for detecting a welded steel pipe bead, a mechanical method, a method using an eddy current sensor, an optical method, and the like have been proposed.
[0004]
[Patent Document 1]
Japanese Patent Publication No.59-2593
[0005]
[Patent Document 2]
JP 2000-176642 A
[0006]
[Patent Document 3]
JP-A-5-133940
[0007]
[Patent Document 4]
Japanese Patent Laid-Open No. 5-18904
[0008]
[Patent Document 5]
Japanese Patent Laid-Open No. 9-72851
[0009]
[Patent Document 6]
JP-A-60-135705
[0010]
As a mechanical method, for example, Patent Document 1 proposes a method of detecting runout of an outer surface welded portion of a traveling pipe with a contact roller.
[0011]
In addition, as a method using an eddy current sensor, in Patent Document 2, a detection head provided with a magnetic core that moves circularly around a concentric axis within a concentric cylindrical transmission coil and a reception coil is welded. Standing above the bead, it is placed in close proximity, and the passage timing of the magnetic core is detected twice from the change in impedance of the receiving coil when the magnetic core passes over the weld bead, and the required time between these passage timings There has been proposed a method for detecting the center position of a weld bead in which the center position of the weld bead is detected by calculating and comparing.
[0012]
In Patent Document 3, a plurality of eddy current sensors are arranged so as to extend substantially in one direction, and the subject is scanned by sequentially switching the roles of excitation, excitation, and detection between the eddy current sensors. A method for detecting a bead position from a detected waveform has been proposed.
[0013]
Further, as an optical method, as disclosed in Patent Document 4, a pipe surface is imaged, and a signal waveform feature quantity specific to a welded part or a base material part is extracted and stored in advance. ITV while irradiating fan-shaped light or irradiating fan-shaped light on the tube surface while rotating the steel pipe in the circumferential direction as disclosed in Patent Document 5 The tube surface is imaged with a camera, etc., and the video signal is corrected to an image that has been corrected for noise removal, inclination, etc., and then the difference between the arc image fitted with the arc by the arc fitting method and the actual image is obtained based on the image. In addition, when the difference data exceeds a preset threshold value, it is determined as a weld bead, and at the same time, the width of the range exceeding the threshold value is encoded as a preset allowable range of the weld bead width. The melt The bead position, the weld bead detection method of the steel pipe have been proposed.
[0014]
In Patent Document 6, as a general technique for automatically measuring the position and shape of a weld bead, the weld bead is imaged from the upper surface and the side surface, and analog image information from the imaging unit is converted into digital information at an ash tone level. An automatic bead shape measuring device that detects the width and height of a weld bead based on the digital image information has been proposed.
[0015]
[Problems to be solved by the invention]
However, in the case of a method using a contact roller as disclosed in Patent Document 1, it is necessary that the height of the bead is substantially constant in the longitudinal direction and the unevenness thereof is relatively steep, and the unevenness of the bead is very smooth. If the bead height is low or the bead height is not constant in the longitudinal direction, accurate detection cannot be performed.
[0016]
In addition, in the case of the method using the concentric circular magnetic core in the concentric cylindrical coil disclosed in Patent Document 2, the seam twisted portion passed through the detection device position or the meandering occurred during the transportation of the electric sewing tube as the subject. In this case, the positional relationship between the electric sewing tube as the subject and the concentric cylindrical coil or the concentric circular magnetic core as the detection device is shifted, so that it is impossible to accurately detect the weld bead position.
[0017]
In addition, in the excitation, excitation, and detection coil substantially one-way multiple arrangement method disclosed in Patent Document 3, it is difficult to avoid erroneous detection because it easily reacts with foreign matters attached to the tube surface or irregularities on the surface other than the beads, Further, when various sizes such as an electric sewing tube are used as a target, a plurality of detection heads must be prepared according to the difference in their shapes, which increases the manufacturing cost of the apparatus.
[0018]
In addition, as a problem common to these eddy current methods, there is a problem in that a separate shape measuring means must be provided in order to evaluate the shape of the weld bead subjected to position detection, and the manufacturing cost of the apparatus increases.
[0019]
Compared to the various methods described above, the optical methods disclosed in Patent Literature 4 and Patent Literature 5 can perform non-contact detection, and not only bead position detection but also the bead shape with the same apparatus configuration. There is an advantage that it can be evaluated. However, the above-described conventional optical method has various problems.
[0020]
In other words, the method based on the feature extraction of the weld base material part disclosed in Patent Document 4 is mainly a method for detecting a difference in brightness between the bead part and other parts (base material part: hereinafter referred to as a raw pipe part). However, since the brightness (brightness) of the bead part greatly depends on the welding conditions and the thickness of the tube, it is difficult to detect stably. In addition, the bead may not be identified particularly if the brightness of the bead part is low. There was a problem.
[0021]
In addition, in the method using the steel tube rotation image arc fitting disclosed in Patent Document 5, it is necessary to rotate the steel pipe in the circumferential direction, but the steel pipe is continuous with the base steel strip at the welding stage of the ERW steel pipe. In addition, there are many problems that cannot be rotated, and since two arcs are calculated from four points of data at the image processing stage, even if noise processing is performed, the jagged shape that is generally found in image data There is also a problem that errors are likely to occur in the calculated bead position, and the roundness of the ERW tube, which is the subject, is often poor, and the tube is placed on two vertical bisectors. This method, which utilizes the geometric property of a circle having a central portion, has a problem in that there is a limit in suppressing the occurrence of detection errors.
[0022]
Further, in the camera imaging method of Patent Document 6, since the point where the image density for one line changes suddenly is searched as a method for specifying the bead position, when the brightness of the bead part is low, There is a problem that the bead position cannot be specified depending on the surface properties.
[0023]
Furthermore, in addition to these detection methods, as described in Patent Document 5, depending on a person skilled in the art, a profile of a steel pipe surface including a bead position is measured by a light cutting method or an optical distance measurement method, and the profile is obtained. A method for detecting the bead position by processing the data is easily recalled. However, in this case, as a method of processing the profile data, it is common to perform some differentiation processing on the assumption that a sudden change occurs in the profile in the bead portion. On the other hand, while the rise is becoming smoother, such differential processing emphasizes minute noise that is likely to occur in optical profile measurement, making it difficult to detect the bead position.
[0024]
The present invention has been made to solve the problems of the prior art as described above. From the shape data of the electric sewing tube by the so-called light cutting method detected by slit light or spot light scanning, the brightness and profile are obtained. It is an object of the present invention to provide a method and apparatus capable of accurately detecting a bead shape without being affected by data noise.
[0025]
[Means for Solving the Problems]
That is, the invention of claim 1 irradiates the welded part of the ERW welded tube with slit light or scans with spot light, and images of the slit light irradiated on the surface of the welded part or the trajectory of the scanned spot light. A bead shape of an electric resistance welded pipe that detects the bead shape of the electric resistance welded pipe by an optical cutting method that performs predetermined image processing on an image obtained by imaging the image from an angle different from the irradiation direction of the slit light. In the detection method, the coordinates of the temporary bead apex are calculated from the profile of the ERW weld pipe by a predetermined calculation formula, the profile of the ERW weld pipe is approximated by a quadratic function to obtain a first approximate curve, Calculate the coordinates of two intersection points between the temporary bead apex of the profile of the ERW weld pipe and the first approximate curve, and calculate the coordinates of the temporary bead apex and the two of the temporary bead apex From the coordinates of the intersection A bead tentative existence range is calculated from the equation, and the shape of the pipe part excluding the bead tentative existence range from the profile of the ERW weld pipe is approximated by an even-order polynomial of second or higher order to obtain a second An approximate curve is obtained, and an area including the coordinates of the provisional bead apex is specified as a bead among areas where the deviation between the profile of the ERW weld pipe and the second approximate curve is greater than a predetermined threshold. A method for detecting the bead shape of an electric resistance welded pipe, characterized by:
According to a second aspect of the present invention, there is provided light projecting means for irradiating slit light or scanning spot light at an angle on a welded portion of an ERW weld pipe, and an image obtained by irradiating the welded portion with the light projecting means. Imaging means for imaging from an angle different from the certain angle, profile calculation means for calculating a profile of the electric resistance welded pipe by performing predetermined image processing on an image obtained by the imaging means, and the electric resistance welding pipe A temporary vertex calculation means for calculating the coordinates of the temporary bead vertex from the profile of the above, and a first regression calculation means for approximating the profile of the ERW weld pipe as a quadratic function by a predetermined regression expression; Intersection calculation means for calculating the coordinates of two intersection points sandwiching the temporary bead vertex from the output of the first regression calculation means and the output of the profile calculation means; the coordinates of the intersection points and the temporary bead A first range calculating means for calculating a provisional existence range of the beads from the coordinates of the points by a predetermined calculation formula; and a range of the electric resistance welded pipe in a range excluding the provisional existence range of the beads calculated as described above. A second regression calculation means for approximating the profile with a second-order or higher-order even polynomial, and a region where the deviation between the output of the second regression calculation means and the profile of the ERW weld pipe is greater than a predetermined threshold value. Of these, a bead shape detection device for an electro-welded pipe, comprising: a second range calculation unit that outputs a value including the coordinates of the provisional bead apex as a weld bead range.
[0026]
DETAILED DESCRIPTION OF THE INVENTION
In the manufacture of ERW pipe, the width of the weld bead before cutting is about 1/10 to 1/5 of the outer diameter of the pipe, and the position of the weld bead can be roughly known in advance. The reason for this is that in the general ERW pipe production line, the metal band that is the material is continuous from the unwinding of the metal band coil such as the steel band that is the base material to the molding and welding of the pipe. This is because the position and orientation are constrained, so that even if horizontal movement of the tube (changes in the pass line), twisting, etc. occur, it is only as high as the bead width. Here, the bead portion may be located at any position on the circumference of the pipe. Of course, no generality is lost by this assumption.
[0027]
Here, the pipe shape including the bead part and sufficiently wider than the bead width (hereinafter referred to as ERW weld pipe profile) is detected by an appropriate method, and the detected ERW weld pipe profile is approximated by a quadratic function. In this case, the approximate curve is a curve that passes above the profile of the raw tube portion and below the apex portion of the bead portion in order to simultaneously approximate not only the raw tube portion but also the shape of the upwardly protruding bead portion. This curve is defined as a first approximate curve. Here, the approximate vertex position Xc of the bead portion is obtained by another method, and the points Xl and Xr at which the first approximate curve intersects the profile of the ERW weld pipe are obtained by searching from the approximate vertex position to the left and right. , Xl, Xr, and Xc, an approximate bead range R (Xl ′ to Xr ′) can be obtained.
[0028]
Here, the first approximate curve is limited to a quadratic function, but this is because the shape of the tube is substantially symmetric with respect to the apex of the bead portion, so that it can be approximated by an even function such as an even-order polynomial. In the meantime, when approximating with a fourth-order or higher order polynomial, an inflection point occurs in the approximate curve and the bead part is emphasized, and this has an unfavorable influence on the calculation of the intersection with the raw pipe part. That is the reason.
[0029]
If the profile in the range excluding R in the width direction coordinate (X coordinate) is approximated by a second-order or higher-order even-order polynomial, the shape of the tube part can be approximated with considerably good accuracy. This is the second approximate curve.
[0030]
The basis of this is the graph of FIG. 9, which shows the RMS of the polynomial and the approximation error when the upper half curve of the ellipse is regressed with a second, fourth, sixth and eighth order polynomial. The average square root) relationship is shown, and it is shown that the shape of an ellipse can be regressed with sufficient accuracy by an even-order polynomial of second or higher order, preferably an even-order polynomial of fourth or higher order.
[0031]
Using this property, the profile of the base pipe part approximated with sufficient accuracy is compared with the profile of the ERW welded pipe including the bead part. A region including the coordinates of can be specified as a bead.
[0032]
In the above calculation, a polynomial is used as the approximate curve. Therefore, if the least square method is used, the regression calculation can be performed only by addition, multiplication, and matrix calculation. In other words, since it does not use any assumption of the perfect circle or differential calculation that was a problem in the prior art, it is not affected by noise, and it also takes time and effort for noise data removal work such as moving average for noise removal and coordinates Such processing is also unnecessary.
[0033]
【Example】
Embodiments of the present invention will be described below with reference to the drawings.
[0034]
FIG. 1 is a schematic view showing a device configuration example of a bead detection device for an electric resistance welded pipe according to the present invention. In FIG. 1, 200 is an electric sewing tube, 10 is a light projecting means, 20 is an imaging means, 30 is a profile calculation device, 40 is a profile data processing device, and 50 is a display device.
[0035]
FIG. 2 is a configuration diagram showing the internal configuration of the profile data processing device 40. In FIG. 2, 100 is a temporary vertex calculation circuit, 110 is a first regression calculation circuit, 120 is an intersection calculation circuit, 130 is a first range calculation circuit, 140 is a second regression calculation circuit, and 150 is a deviation calculation circuit, Reference numeral 160 denotes a second range calculation circuit.
[0036]
In FIG. 1, the light projecting means 10 is a slit light source in which light emitted from a light emitting element such as a laser or a lamp is linearly converged by a cylindrical lens or the like, or light that converges in a point shape at an irradiation position is mirrored. However, it is desirable to use a small slit light source that integrates a semiconductor light emitting element (LED) and a lens system, and the short side width of the slit is also welded. It is desirable that the height be sufficiently smaller than the height of the bead, and it is desirable that it is 50 μm or less. Ultimately, the shape of the part to be measured is calculated as a single line by light-cutting image processing, which will be described later, so this is not essential, but it should be as small as possible.
[0037]
As the imaging means 20, an ITV camera or PSD (optical position detection element) can be used, but a CCD camera is preferably used in consideration of ease of data conversion to a subsequent image processing apparatus. It is. Although not shown in FIG. 1, a lens mechanism for forming an image of irradiation light, a diaphragm and a shutter mechanism for adjusting the amount of received light to an appropriate range, and the like can be generally selected and mounted. Good. Here, when a method of scanning a point light source as the light source is adopted, it goes without saying that it is necessary to continue irradiation while scanning the entire range in the width direction at least once. If the tube and bead shape do not change while satisfying this condition and scanning is completed, the acquired image will be the same for both slit light and point light source plane scanning. I will explain only.
[0038]
The incident angle α from the light projecting means 10 that is a light source and the arrangement angle of the imaging means 20, that is, the imaging angle β is preferably (α + β) approximately 90 °, and the number of pixels and the field of view of the camera that is the imaging means 20 are The width may be determined based on the width of the bead portion and the necessary resolution. In the present invention, the slit light irradiation angle α = 60 ° from the light source, the imaging angle β = 30 °, the field of view is width (horizontal) × height (vertical) = (25 mm × 20 mm), and the number of pixels is horizontal × vertical = 640 × 480 pixels were used as the preferred value. As a result, the resolution in the width direction is
25/640 = 0.0391 (mm)
The resolution in the height direction is
20/480 * cos (60 °) / sin (60 ° + 30 °) = 0.0209 (mm). In this embodiment, the resolution in the width direction (tube circumferential direction) is 40 μm and the height direction (tube axis direction) is 20 μm. The bead shape can be detected.
[0039]
The profile calculation device 30 converts an image of slit light reflected on the tube surface as illustrated in FIG. 3 into a single line by an appropriate image processing means, and further, from the arrangement of the light source and the imaging device, that is, the irradiation angle α and the like. The geometrical calculation is performed from the imaging angle β to calculate the image of the slit light, that is, the pseudo cross-sectional profile, the true profile in the thickness cross-sectional direction of the pipe, that is, the profile data of the ERW welded pipe. Here, as the image processing means, it is generally possible to use an apparatus that performs thinning processing, but preferably the thinning processing means previously proposed by the inventors in Japanese Patent Application No. 2002-128497 (not disclosed at the time of filing) is used. Is good. In addition, for detecting the bead shape, which is the object of the present invention, the profile data of the pipe is simple, so that the step of geometric calculation described above may be omitted, so that another step can be supported.
[0040]
Next, each part in the profile data processing apparatus 40 will be described below. In the pseudo cross-sectional profile formed by slit light irradiation as described above, or the profile of the ERW welded pipe (both are assumed to be after the tapering process), X in the direction (width direction) crossing the bead portion. Taking the axis, the profile can be represented as a data group of height corresponding to the X coordinate.
[0041]
Here, the provisional vertex calculation circuit 100 calculates the weld bead vertex position Xc0, and may be configured to calculate, for example, the load average (center of gravity position) of profile data.
[0042]
For a certain width direction (X-axis direction) coordinate, the value obtained by multiplying the luminance and the vertical axis coordinate of the pixel indicating it is added in the vertical axis direction, and the average value divided by the number of participating pixels is Obtain the average value for other width direction (X-axis direction) coordinates in the same way, and connect the average values in the width direction (X-axis direction). The X coordinate shown is to be obtained.
[0043]
Or, more simply, the pixels indicating the maximum luminance with respect to a certain width direction (X-axis direction) coordinate are connected in the width direction (X-axis direction), and the X coordinate indicating the maximum value of the vertical axis is obtained from them. It may be.
[0044]
Further, since the first regression calculation circuit 110 regresses profile data as shown in FIG. 4 with a quadratic function, the first regression calculation circuit 110 is configured to execute a known regression calculation, preferably a least square calculation rule. do it. A first approximate curve obtained by approximating the profile data with a quadratic function by the first regression calculation circuit 110 is as shown in FIG.
[0045]
Here, in FIG. 4, a relative value is used as the height of the bead indicated by the vertical axis. As described above, this relative value is easy for the profile data of the ERW welded pipe for the purpose of detecting the bead shape, which is the object of the present invention. Is a value obtained by omitting the geometric calculation part (hereinafter the same).
[0046]
As shown in FIG. 5, the intersection calculation circuit 120 selects the two closest points on the left and right sides of the vertex Xc0 from the points where the profile data and the first approximate curve intersect, and sets them as Xl and Xr, respectively.
[0047]
The first range calculation circuit 130 calculates a range of coordinates used for the following tube segment regression based on Xl and Xr calculated by the intersection calculation circuit 120 and the vertex value Xc0. For example, the 3: 2 outer dividing point between the vertex and the intersection
Xl '= (3X1-Xc0) / 2
Xr '= (3Xr-Xc0) / 2
R: x <Xl ′, x> Xr ′
The external ratio may be determined appropriately in consideration of, for example, the average smoothness of the bead rise.
[0048]
The second regression calculation circuit 140 may be configured to perform the same least square polynomial regression calculation as the first regression calculation circuit in the range R of x calculated as described above. However, in the second regression calculation circuit 140, the order of the polynomial to be calculated is an even-order polynomial of 2nd order or higher, and is preferably configured to be a polynomial of 4th order or higher. A second approximate curve is thus obtained. By the way, this second approximated curve extends complementarily to the region of Xl ′ ≦ x ≦ Xr ′ that deviates from the range R of x.
[0049]
The deviation calculation circuit 150 is the second regression calculation circuit described above over the entire X coordinate where the profile data of the ERW welded pipe exists (in the sense that it can be captured in the image field as a light cut image). The deviation between the second approximated curve output from the profile data of the ERW welded pipe and the profile data of the ERW welded pipe is calculated, and can be constituted by a polynomial arithmetic circuit and a subtraction circuit.
[0050]
The second range calculation circuit 160 calculates a range in which the output of the deviation calculation circuit 150 exceeds a predetermined threshold, and outputs a portion including the vertex Xc0 in this range as a temporary existence range of beads. Thus, a threshold circuit and a comparator circuit can be used.
[0051]
Hereinafter, the operation of the present embodiment will be described using data.
[0052]
The light section image irradiated with slit light from the light projecting means 10 and imaged by the imaging device 20 is as shown in FIG. 3, and the bead portion calculated by the profile calculation device 30 is included for this light section image. FIG. 4 shows a thinning process of profile data of an electric resistance welded pipe. The temporary vertex calculation circuit 100 calculates a vertex for the profile data by a technique such as maximum value calculation or weighted average (centroid calculation). Xc0 entered in FIG. 4 is the position of the apex of the bead calculated in this way.
[0053]
The first regression calculation circuit 110 performs a least square regression calculation based on a quadratic expression of the entire profile, and as a result, a quadratic function as shown in FIG. 5 is output. Similarly, Xl, Xr, and Xl entered in FIG. ', Xr' is the position of the X coordinate calculated by the intersection calculation circuit 120 and the range calculation circuit 130 as described above.
[0054]
The second regression calculation circuit 140 calculates a second approximate curve of profile data for the range of the X coordinate set by the range calculation circuit 130. In this embodiment, it is assumed that the fourth order regression is performed as a suitable example of the regression order. The second approximate curve obtained as a result is as shown by the thick line in FIG.
[0055]
The deviation calculation circuit 150 calculates the deviation e (x) between the thick line in FIG. 6 and the profile data of the ERW welded pipe, and obtains the result as shown in FIG.
[0056]
Then, the second range calculation circuit 160 searches for a range in which the deviation e (x) exceeds a preset threshold value, and calculates an X coordinate range including the vertex coordinate Xc0. In this embodiment, 0.05 is used as a preferred example of the threshold value. The result shows the range of the arrow entered in FIG.
[0057]
In order to confirm the validity of the present invention, a photograph of a bead photographed by stopping the light emission of the light source and extending the exposure time in the same arrangement of the light projection means and the image pickup means as the apparatus of the present embodiment, and the bead according to the present invention I decided to compare the image output of shapes. Although the result is as shown in FIG. 8, it can be seen that the result is in good agreement with the calculated bead shape (shown in the lower part of FIG. 8).
[0058]
In the embodiment described above, it is natural that some or all of the internal configuration circuits of the profile calculation device 30 and the profile data processing device 40 may be realized as software in a digital computer or ROMized program. Of course, the present invention is applicable not only to steel pipes but also to copper, aluminum, and other metal pipes.
[0059]
【The invention's effect】
As described above, according to the present invention, an electro-welded pipe that is detected by a light cutting method using slit light projection or point light source scanning without performing a large operation such as rotation of the pipe or complicated calculation processing such as differentiation. Since the bead shape is detected from the shape data by the regression calculation that does not involve the assumption of the shape or the differential operation, etc., even when the brightness of the bead changes or the roundness of the blank tube part decreases The bead shape can be reliably detected. Further, since the bead shape is detected only from the non-contact optical detection means and the data calculation means, detection with high objectivity and reproducibility becomes possible. In addition to simply displaying the bead shape, it can be used for advanced operations such as continuously recording the trend, and using the specified bead shape information for more advanced bead shape analysis. It has an excellent effect.
[Brief description of the drawings]
FIG. 1 is a schematic diagram showing a configuration of a main part of a weld bead detection device for an electric resistance welded pipe according to the present invention.
FIG. 2 is a block diagram showing the configuration of an arithmetic circuit group that also constitutes the profile data processing apparatus.
FIG. 3 is a view showing an example of a light section image near the bead of the ERW tube
FIG. 4 is a view showing profile data obtained by thinning the light section image near the bead of the ERW tube.
FIG. 5 is a diagram showing a state of a first approximate curve of a quadratic function calculated by regression calculation for the entire profile data, similarly output from the first regression calculation circuit;
FIG. 6 is a diagram showing a state of a second approximate curve of a quaternary function obtained as a result of the least square regression calculation with respect to the range of the first range calculation circuit, similarly calculated by the second regression calculation circuit;
FIG. 7 is a plot of deviation e (x) between the quaternary function and profile data calculated by the deviation calculating circuit.
FIG. 8 is a diagram showing a comparison between the present embodiment and a photograph in which the actual bead is photographed with the same optical system arrangement;
FIG. 9 is a diagram showing the relationship between the degree of polynomial and the RMS (root mean square) of approximation error when the upper half curve of the ellipse is regressed with a second-order, fourth-order, sixth-order, and eighth-order polynomial;
[Explanation of symbols]
10 ... Projection means
20: Imaging means
30 ... Profile calculation device
40 ... Profile data processing device
100: Temporary vertex calculation circuit
110: First regression calculation circuit
120 ... Intersection calculation circuit
130... First range calculation circuit
140: Second regression operation circuit
150: Deviation calculation circuit
160... Second range calculation circuit
200: Electric pipe
210 ... Bead

Claims (2)

Irradiating the welded portion of the ERW welded tube with slit light or scanning with spot light, and the slit light image or the scanned spot light trajectory image on the surface of the welded portion is defined as the irradiation direction of the slit light. In the bead shape detection method for an electric resistance welded pipe, the bead shape of the electric resistance welded pipe is detected by an optical cutting method in which predetermined image processing is performed on an image obtained by imaging with imaging means from different angles.
Calculate the coordinates of the temporary bead apex from the profile of the ERW welded pipe using a predetermined formula.
A first approximate curve is obtained by approximating the profile of the ERW weld pipe with a quadratic function,
Calculating the coordinates of two intersections of the ERW welded pipe profile and the first approximate curve with the provisional bead apex sandwiched between them,
From the coordinates of the temporary bead vertices and the coordinates of two intersection points sandwiching the temporary bead vertices, a temporary existence range of the beads is calculated by a predetermined calculation formula,
Approximating the shape of the pipe part excluding the temporary existence range of the bead from the profile of the electric resistance welded pipe with a second-order or higher-order even-order polynomial to obtain a second approximate curve,
Among the regions where the deviation between the profile of the ERW weld pipe and the second approximate curve is greater than a predetermined threshold, the region including the coordinates of the provisional bead apex is specified as a bead, A method for detecting the bead shape of an ERW welded pipe. A light projecting means for irradiating slit light or scanning spot light at an angle in the welded portion of the ERW pipe;
An imaging unit that captures an image of the light projecting unit irradiated on the welded portion from an angle different from the certain angle;
Profile calculating means for calculating a profile of the electric resistance welded pipe by applying predetermined image processing to an image obtained by the imaging means;
Provisional vertex calculation means for calculating the coordinates of the provisional bead vertex from the profile of the ERW weld pipe by a predetermined calculation formula;
First regression calculation means for approximating the profile of the ERW weld pipe as a quadratic function by a predetermined regression equation;
An intersection calculation means for calculating coordinates of two intersections sandwiching the temporary bead vertex from the output of the first regression calculation means and the output of the profile calculation means;
A first range calculating means for calculating a temporary existence range of beads by a predetermined calculation formula from the coordinates of the intersection and the coordinates of the temporary bead vertex;
A second regression calculation means for approximating the profile of the electric resistance welded pipe in a range excluding the temporary existence range of the beads calculated as described above by an even-order polynomial of second order or higher;
Among the regions where the deviation between the output of the second regression calculation means and the profile of the ERW weld pipe is greater than a predetermined threshold, the one including the coordinates of the temporary bead apex is output as the weld bead range. Second range calculating means for
An apparatus for detecting a bead shape of an electric resistance welded pipe, comprising:

Images