JPH0332470A - Automatic shape recognizing method for u-shaped groove - Google Patents

Abstract

PURPOSE:To accurately detect groove shape elements by calculating a groove width, groove depths, difference in level, stagger and a root gap based on four characteristic points. CONSTITUTION:The picture data of a U-shaped groove shape obtained by the optical cutting method are converted into a two-dimensional X-Y coordinate system to calculate the groove shape elements by using a computer. Points P1 and P2 where an error between a measured value and a reference value is made more than (l) dot are then obtained from right and left inflection points. Intersections P3 and P4 between regression equation of respective right and left shoulder heights and regression equation of rectilinear parts of slants are obtained by the expressions I and II, respectively. These P1, P2, P3 and P4 are made to the characteristic points and based on those characteristic points, the groove width W, the groove depths H1 and H2, the difference in level DELTAh1, the stagger DELTAh2 and the root gap DELTAw are calculated. By this method, confirmation and record holding of the groove shape elements having objectivity and reliability can be carried out.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an automatic shape recognition method for a U-shaped groove in which a two-dimensional graphical analysis of the U-shaped groove is performed using a computer.

[Conventional technology]

Conventionally, for quality control of welded products that are butt welded using a U-shaped groove, groove width W, groove depth H2, step difference Δh29, discrepancy Δh2 are required for the U-shaped groove l shown in the explanatory diagram in Fig. 5.
．． Bevel shape elements such as root gasop ΔW are measured according to the procedure shown in the flowchart in Figure 6, and inspected to confirm whether the design tolerances are satisfied.
As shown in the explanatory diagram in Figure (A), the inspection gauge 10 is used to determine whether the maximum and minimum range of design tolerances are satisfied. ) As shown in the explanatory diagram, shape elements such as the groove width W to the root gap ΔW of the portion whose detailed data is to be measured are measured using a measuring gauge 11.

However, such conventional inspection methods have the following drawbacks.

(1) Since a gauge is used for measurement, the test results vary depending on the skill level of the inspector, resulting in a lack of objectivity and reliability.

(2) Inspection records include recording of on-site measurement data by inspectors;
This was done through rearrangement, resulting in a significant lack of recording reproducibility.

(3) Unable to respond to demands for rationalization of inspections and labor savings.

[Problem to be solved by the invention] The present invention was proposed in view of the above circumstances, and
The groove shape elements can be accurately detected using a computer from the shape image data of the U-shaped groove obtained by the optical cutting method, and the groove shape elements can be confirmed and recorded with objectivity and reliability. An object of the present invention is to provide an automatic shape recognition method for a U-shaped groove that can be maintained.

[Means to solve the problem]

To this end, the present invention uses a computer to convert image data of a U-shaped groove shape obtained by the optical cutting method into a two-dimensional X-Y coordinate system, and calculates groove width and groove depth as groove shape elements.

In calculating and detecting the difference in level difference and root gap, (1) For the height of the left and right shoulders of the groove in the two-dimensional X-Y coordinate system, take the average value of the data for each shoulder, and the error from that is large. The shoulder height is the average value of the data up to n points before the inflection point, and the circular orbit formula y = aVr, +y
= avr2, (2) Next, for the straight line parts of the left and right slopes of the groove, calculate the rate of change dy/d m times from the data n points ahead of the inflection point, and use the average value as the slope. 2, the circular orbit formula in the X-Y coordinate system)' = k I X
+ b 1 , y − is expressed as 2 x + b2,
(3) Furthermore, the error between the measured value and the reference value is l from the left and right inflection points.
Point Pl (Xt * '11) + P that is more than a dot
2 (X2 * y2), and the intersection point Ps (Xs, ys) of the circular orbit formula at the height of the left and right shoulders and the circular orbit formula at the straight part of the slope: x5 = (av
rlbt ) /kt + P+ (Xs, 3
F! ): Find 2 in Xs= (avr2-b2) / and calculate these P, r P 2 , P 3 *
P4 is a feature point, and (4) the above feature point P1. P2
, Ps, P+, groove width W''
s * Groove length H1=y1731 H2=3'2)
'41 step difference ΔKt = 1+-y51 discrepancy Δh2"
)'2)'s + root gap Δw=x2-Xt
It is characterized by calculating.

[Effect]

In the method of the present invention, a U-shaped groove is irradiated with band-shaped slit light, image data of the groove shape taken with a television camera is recorded in an image memory, and a computer is used to create a two-dimensional X-Y coordinate system. Replaced with the height a of the shoulders on the left and right sides of the groove.
Calculation of vrl, avr2 - slope dy of left and right slopes
Calculation of /lx - Point traced from the inflection point, intersection point P of the left and right shoulder heights and the left and right slopes.

P2. Calculation of P, , P and - Seki width W, groove depth H, l
H21 step difference Δh29 discrepancy Δh2゜ route gap Δ
Calculations are performed in the order of calculation of W, and the shape of the U-shaped groove is automatically detected and recognized.

〔Example〕

One embodiment of the automatic shape recognition method for a U-shaped groove according to the present invention will be explained with reference to the drawings. Figure 1 is a perspective view showing the implementation device, Figure 2 is an explanatory diagram of groove shape elements, and Figure 3 is a groove diagram. Flowchart of shape recognition process, FIG. 4 is an explanatory diagram of each process of groove shape recognition.

First, FIG. 1 shows a structure in which a U-shaped groove l is measured by the optical cutting method, and the image data of the groove shape recorded in the image memory 8 as array data is calculated by the computer 9 as groove shape elements. , shows an aspect, and the photocleavage method is U
A slit light 4 emitted from a laser light source 2 and a cylindrical lens 3 is irradiated onto the shaped groove l as a band-shaped slit light 6 via a
When observed with
This is a method that allows you to measure the shape as if it had been cut.

Next, the outline of the automatic groove shape recognition method in the computer 9 will be explained below with reference to FIGS. 2 to 4.

That is, the groove shape elements of the U-shaped groove 1 as shown in FIG. 2 are groove width W, groove depth Hl + H2 * step difference Δ
In order to obtain the h19 discrepancy Δh2° root gap ΔW, calculations are performed using the process shown in the flowchart of FIG.

First, data subjected to geometric correction to remove distortion of the figure at the time of observation is stored in, for example, 502 arrays and replaced with an X-Y coordinate system.

The method for determining the shoulder height is as shown in Fig. 4 (1), if the error between the geometric correction data and the design value reference data is, for example, ±10.
While checking that it is within the dots, take the difference from the next data.The point where this difference is, for example, 25 dots or more is the inflection point A, A-10 is m, and 1 is, for example, 5.
0, and the average of the data from l to m is the height of the left shoulder. Similarly, for the height of the right shoulder, look at the data in the negative direction, and define the point where the difference is 25 dos or more as the inflection point B.
+10 is m, 1 is, for example, 480, and the average value of the data from l to m is the height of the right shoulder.

Then, the height of the left shoulder avrl is determined by equation (1).

Find the height avr2 of the right shoulder.

To find the slope of the straight line part of the slope, in Fig. 4 (II), while confirming that the error between the geometric correction data from A+5 and the reference data is within, for example, ±10 dots, set dy/dx, for example. Obtain it 5 times, use the average value as the left slope, and similarly look at the data in the negative direction from B-5, d
Determine y/dχ five times and use the average value as the right slope.

Then, according to equation (2), the left slope is 1. 2 on the right slope
seek.

k under snow Σdy/dx ... (2) Feature point Ps
+ P 21 Ps r P + is found in the fourth
In figure (III), the straight line Ltd)'s=k, x+b
, and y”avrl is the intersection point P.

(Xs + 3'), calculated using equation (3).

! ! = (avrl-bl)/fk
.

...(3) Straight line L2: y2 = k2x + b2 and y = avrl
The intersection point with PG (x, y4) is determined by equation (4).

x, = (avr2 b2)/fk2... (4) If the error between the geometric correction data and the reference data is taken in the positive direction from X, and there are 5 consecutive points where this error is 110171 or more The first point is P, (Xt+
yt).

Similarly, take the error between the geometric correction data and the reference data in the negative direction from x4, and if there are 5 consecutive points where this error is ±10 dots or more, the first point is set to F2.
(X21)F2).

Thus, based on the above calculated values, in Fig. 4 (EV) and (V), the groove width W and the groove depth H are
l + 82 + step difference Δh19 discrepancy Δh2 and root gap ΔW are determined as follows.

The groove width W is determined by equation (5).

W=x, -Xj (5) The left groove depth Hl and the right groove depth H2 are determined by equation (6).

Hl = hosbuf (xl) - aVFl, H2
=hosbuf (x2) -avr2 (6) The step difference Δh is determined by the equation (7).

Δh,=H2-H,...(7) The discrepancy Δh is determined by equation (8).

Δh7 =hosbuf (x2) −hosbuf
(Xt)...(8) Find the root gap ΔW using equation (9).

Δw-x2-Xl . . . (9) As described above, the groove shape elements of the U-shaped groove 1 can be automatically and accurately calculated and recognized.

〔Effect of the invention〕

In short, according to the present invention, U obtained by photosection method
Image data of the groove shape is converted into a two-dimensional X-Y coordinate system using a computer.In calculating and detecting the groove width, groove depth, one step difference, and the root gear/knob as the groove shape elements, ( 1) The height of the shoulders on the left and right sides of the groove in the two-dimensional X-Y coordinate system is determined by taking the average value of the data at each shoulder, and then averaging the data up to n points before the inflection point where the error from that value becomes large. The value is the height of the shoulder, and the orbital formula y
= aVrj, y'' expressed as aVr7, (2) Next, the straight line portions of the left and right slopes of the groove are n from the inflection point, respectively.
The rate of change dV/dx@ m times is calculated from the previous data and the average value is set as the slope 2. 2, the circular orbit formula y=k, x+b, in the X-Y coordinate system.

y=k2) c+b2, and (3) point P where the error between the measured value and the reference value is more than β dots from the left and right inflection points.
s (X+, Y+), P2' (x2゜y2), and find the intersection point PI (Xj +
3'3): Xj = -(aVrlbl)/k
t, PI (X4 + '!+): X4=
(avr2 b2 ) Find 2 in / and add these Pl
, p2. Let p,,P+ be the feature points, (4) the above feature point P
1. Groove width W”X based on P2, Pl, P+
4-X g + groove depth H1 = y.

y3・H2=y2)'Now 9 steps Δh,"')'+
-y31 discrepancy Δh2 = y2 Ys, by calculating the root gap ΔW” The present invention is industrially extremely useful because it provides an automatic shape recognition method for a U-shaped groove that can objectively and reliably confirm groove shape elements and maintain records thereof. .

[Brief explanation of drawings]

1 to 4 show an embodiment of the automatic shape recognition method for a U-shaped groove according to the present invention.
FIG. 3 is an explanatory diagram of groove shape elements, FIG. 3 is a flowchart of the groove shape recognition process, and FIG. 4 is an explanatory diagram of each process of groove shape recognition. Figures 5 to 7 show the conventional U-shaped groove inspection method.
FIG. 6 is an explanatory diagram of groove shape elements, FIG. 6 is a flowchart of the groove shape inspection process, and FIG. 7 is an explanatory diagram of each inspection aspect of the groove shape inspection. 1...U-shaped groove, W...groove width, Hl, H2...
Groove depth, Δh,... step, Δh2... discrepancy,
ΔW... Root gap, 2... Laser light source, 3
... Cylindrical lens, 4... Slit light, 5
...mirror, 6...belt-shaped slit light, 7...television camera, 8...image memory, 9...computer

Claims (1)

[Claims] The image data of the U-shaped groove shape obtained by the optical cutting method is converted into a two-dimensional X-Y coordinate system using a computer, and the groove shape elements include groove width, groove depth, When calculating and detecting steps, discrepancies, and route gaps, (1)
The height of the left and right shoulders of the groove in the two-dimensional X-Y coordinate system is determined by taking the average value of the data at each shoulder, and then calculating the average value of the data up to n points before the inflection point where the error becomes large. As the height of the shoulder, the orbital formula y=avr_1, y=avr
(2) Next, for the straight line parts of the left and right slopes of the groove, calculate the rate of change dy/dx m times from the data n points ahead of the inflection point, and calculate the average value as the slope k_1, k_2.
As, the circular orbit formula y=k_1x+b in the X-Y coordinate system
_1, y=k_2x+b_2, (3) Furthermore, points P_1 (x_1, y_1), P_2 (x_2, y_
2) and the intersection point P_3(x_3,
y_3):x_3=(avr_1-b_1)/k_1,
P_4(x_4, y_4):x_4=(avr_2-b
_2)/k_2 and calculate these P_1, P_2, P_
3. P_4 is a feature point, (4) the above feature points P_1, P
Based on _2, P_3, and P_4, groove width W=X_4-X_
3. Groove depth H_1=y_1-y_3, H_2=y_2
-y_4, step difference Δh_1=y_4-y_3, discrepancy Δ
h_2=y_2-y_1, root gap Δw=x_2
An automatic shape recognition method for a U-shaped groove, characterized by calculating -x_1.