JP3102984B2 - Groove position detection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a groove position detecting method for photographing a welding groove with a television camera and using an image signal from the television camera to control an automatic welding machine.

[0002]

2. Description of the Related Art A groove position detecting method for following and controlling an automatic welding machine using an image signal of a television camera is disclosed in Japanese Patent Application Laid-Open No. 52-85044 or Japanese Patent Application Laid-Open No. 53-9235.
As disclosed in Japanese Patent Application Publication No. 3 (1999), there has been proposed a method for detecting a groove by using a difference between the lightness of a groove illuminated by a spot light and the surface of a welded plate. The image signal of the television camera has a high voltage in a white (light) portion where the brightness of the subject is high. Further, the voltage is low in a black portion (dark portion) where the luminance is low. As shown in FIG. 3A and FIG. 3B, the image signal voltage changes in accordance with the brightness (shading). In the image of the groove and its surroundings, since the inside of the groove is dark (dark) and the surface of the weld plate is light (bright), binarization is performed using an appropriate voltage threshold, and the low voltage portion is grooved and the high voltage portion is used. Can be determined as the surface of the welded plate. JP-A-52-85044 and JP-A-5-85044
In Japanese Patent Application Laid-Open No. 3-92353, an area illuminated by a spotlight is binarized as a pattern to be detected, and the spotlight and the welding torch are moved and copied so that a low voltage portion corresponding to a groove is at the center of the pattern. It is said that it can be controlled.

The purpose of binarization is one of means for separating only the target image from the background, and is effective when the luminance of the target image is different from the background. However, the surface of the weld plate has rust, paint, grinder polishing, tack weld bead, spatter adhesion, etc., and the inner surface of the groove is not always constant brightness due to grinder polishing, tack weld bead, spatter adhesion, etc. . Therefore, when these brightness fluctuations are included in the pattern to be detected illuminated by the spot light, the groove cannot always be properly separated and determined in the binarized image, and the stable groove is not obtained. Is difficult to detect.

[0004]

The problem with these conventional groove position detection methods is that the brightness of rust, paint, grinder polishing, tack weld bead, spatter adherence, etc., causes fluctuations in brightness. Binary processing has been a problem as a method of separating and extracting the tip from the background. SUMMARY OF THE INVENTION It is an object of the present invention to provide a groove detection method that applies a method other than the binarization process as a method of determining a groove, removes the influence of disturbance, and performs stable scanning control.

[0005]

According to the present invention, there is provided a method for detecting a position of a welding groove by using an image signal voltage of a television camera. The gray image in the vicinity of the tip is converted from analog to digital, and the gray image data of s × t pixels in the vicinity of the groove before the scanning control consisting of a two-dimensional image signal voltage is stored. Nx > m pixels, s> n,
The gradation image data in the region of t> m is stored as a teaching pattern, the gradation image photographed in the vicinity of the groove taken during the contour control is converted from analog to digital, and the vicinity of the groove in the contour control consisting of a two-dimensional image signal voltage is converted. S × t pixels of density image data, and n × m pixels of the same size as the teaching pattern , s
> N, t> m are sequentially taken out as an input pattern,
Laterally i-th from the top left of the teaching pattern, longitudinally j th
The brightness W (i, j) at the eye position and whether
From the i-th position in the horizontal direction and the j-th position in the vertical direction
j) is calculated by the following equation (6). The area where the correlation coefficient R is 0.6 to 1.0 and the correlation coefficient is the largest is defined as the groove position.

(2) The grayscale image in the groove width direction near the groove taken before the scanning control is converted from analog to digital, and the grayscale image data in the groove width direction before the scanning control consisting of one-dimensional image signal voltage is stored. to the vertical direction from the upper left corner of the gray-scale image data
From the left end consisting of the groove and the surface of the weld plate from the luminance distribution C (x) in the y-th horizontal direction (x-direction) from any q-th
The grayscale image data of the section up to q + mth is the teaching pattern
Stored as D (i) , near the groove taken during profiling control
The dark light image by analog / digital converter, storing the gray-scale image data, from among the light and shade image data in the control had該倣
Any q numbered from the left end of the teaching pattern equal size
Eyes and sequentially extracts the input pattern section up q + m-th from the teaching pattern D (i) and the correlation coefficient R of the image signal voltage of the input pattern below (10) The area where the correlation coefficient R is 0.6 to 1.0 and the correlation coefficient is the largest is defined as the groove position.

(3) In the groove position detecting method according to the above (1) or (2), the grayscale image data of the region consisting of the groove and the surface of the welding plate is selected from the grayscale image data near the groove before the scanning control. Is performed for each groove having a different gap, one teaching pattern of each groove having a different gap is stored, and the correlation between the teaching pattern and the image signal voltage of the input pattern is calculated. The process of obtaining the area having the largest correlation coefficient in the range of 0.6 to 1.0 is executed a plurality of times by the number of teaching patterns, and the area having the largest correlation coefficient is set as the groove position.

(4) In the groove position detection method according to the above (1), a process of storing, as a teaching pattern, grayscale image data of a region including a groove and a weld plate surface from among the grayscale image data near a groove before scanning control. Is performed for each groove having a different line scanning deviation amount or a line scanning deviation amount and a gap, and teaching patterns of the groove having a different line scanning deviation amount or a line scanning deviation amount and a gap are stored one by one. The process of calculating the correlation of the image signal voltage of the input pattern and executing the process of obtaining the region having the correlation coefficient of 0.6 to 1.0 and having the largest correlation coefficient a plurality of times for the number of teaching patterns is performed, and the region having the largest correlation coefficient is opened. Set to the first position.

(5) In the groove position detection method according to the above (4), the same plane as the shaded area in the vicinity of the groove during scanning control, which is composed of t × s pixels of horizontal t pixels and vertical (running direction) s pixels. A reference point O is provided in advance at one point above, and the coordinate position of the groove position detected when the pixel is converted from the reference point O to (t1, s
1) The coordinate position of the control point P before correction is (t2, s2),
The linear scanning deviation amount of the teaching pattern having the largest correlation coefficient is represented by T
Then, the coordinate position (t3, s2) of the groove position at the corrected control point P is set to t3 = t1− (s2−s1) × T.

(6) In the groove position detecting method according to the above (1), (2), (3) or (4), the two plates constituting the groove of the teaching pattern stored before the scanning control are used. Assuming that the average value of the digital values of the gray image on the surface after analog / digital conversion is D1 and D2 and the digital value of the gray image in the groove after analog / digital conversion is Db, the following (1),
A pseudo groove drawn so as to satisfy the equations (2) and (3) is photographed, the grayscale image is converted from analog to digital, and the grayscale image data near the groove before scanning control consisting of a two-dimensional or one-dimensional image signal voltage is obtained. The gradation image data of the pseudo groove portion is stored as a teaching pattern from among the gradation image data near the groove before the contour control, or the pseudo groove satisfies the expressions (1), (2) and (3). Pseudo-grooves satisfying the formulas (2), (3) and (4) or a plurality of pseudo-grooves drawn so as to satisfy the formulas (2), (3) and (5) are respectively photographed, and the grayscale image is converted into an analog / digital image. Digitally converting and storing gray-scale image data near a groove before scanning control consisting of a two-dimensional or one-dimensional image signal voltage, and gray-scale image data of a pseudo groove portion from the gray-scale image data near the groove before scanning control as a teaching pattern. When -0.16 ≦ (D1−D2) / (Dmax−Dmin) ≦ 0.16 (1) Dmin <Db <D1 <Dmax (2) Dmin <Db <D2 <Dmax ... (3) 0.16 <(D1-D2) / (Dmax-Dmin) <1.00 ... (4) -1.00 <(D1-D2) / (Dmax-Dmin) ) <-0.16 (5) where Dmax is the maximum value of the digital value that can be obtained by converting the image signal from analog to digital. Dmin: is the minimum value of the digital value that can be obtained by converting the image signal from analog to digital. (7) In the groove position detecting method according to the above (1), (2), (3) or (4), digital values indicating the grayscale images of the surfaces of the two plates constituting the groove are D1 and D2. Assuming that the digital value indicating the grayscale image in the groove is Db, the digital value satisfying the above formulas (1), (2) and (3) is used as the grayscale image data of the pseudo groove portion and the groove of the teaching pattern D1 and D2 are written in the memory addresses corresponding to the surfaces of the two plates constituting D, and Db is written and stored in the memory addresses corresponding to the grooves, or (1),
Digital values satisfying equations (2) and (3), (2),
A digital value satisfying the expressions (3) and (4) or a digital value satisfying the expressions (2), (3) and (5) is
D1 and D2 are stored in the memory addresses corresponding to the surfaces of the two plates forming the groove of each teaching pattern as gray image data of a plurality of pseudo groove portions, and Db is stored in the memory address corresponding to the inside of the groove. Is written and stored. Here, the vicinity of the groove is, specifically, about 5 to 17 around the groove.
5mm square. In addition, the line displacement of the groove in the groove width direction is a deviation amount of the groove in the width direction with respect to the traveling direction, and is defined as a change amount of the groove position in the groove width direction with respect to the unit traveling distance as follows: Copying deviation = distance of change in groove position in groove width direction / unit traveling distance The groove position is the center of the groove in the width direction.

[0011]

Hereinafter, the present invention will be described in detail. Generally, an image signal of a television camera is output as an analog voltage corresponding to the luminance, but the image signal output order is such that horizontal luminance signals from the left end to the right end of the screen are sequentially arranged from the top to the bottom in the vertical direction of the screen. Is output. Therefore, in order to convert the image signal from analog to digital and handle it as image data, the number of pixels in the horizontal direction and the number of pixels in the vertical direction are determined in advance, and the number of memories corresponding to the total number of pixels is prepared. The analog voltage of the luminance signal output in the above order is converted into digital data while taking a timing so that the luminance at an arbitrary position is associated with the data at the predetermined address in the memory, and the predetermined voltage in the memory is obtained. To store the data. Images taken before the copying control and during the copying control are subjected to analog / digital conversion of the image signal and stored in a memory composed of s × t pixels (s and t are positive integers).

First, in the two-dimensional groove position detecting method described in the above (1) (corresponding to claim 1), a groove image (FIG. 2) is photographed in the vertical direction before scanning control and stored in a memory. From inside
N × m pixels (n, n) including the groove 21 and the weld plate surface 22
m is a positive integer, and stores a two-dimensional teaching pattern 50 ((a) in FIGS. 2 and 3) of a grayscale image composed of s> n, t> m).

The sputter 23 and the paint character 2 are included in the vertical groove 21 and the peripheral image taken during the copying control.
FIG. 1 shows an image composed of vertical and horizontal s × t pixels in which the tack welding 25 is photographed. FIG. 4A shows an input pattern 51 near spatter and FIG. 5 shows an input pattern 52 near paint characters of the same two-dimensional size of n × m pixels as the teaching pattern 50 in FIG. FIG. 6A shows an input pattern 53 near the tack welding and FIG. 7A shows an input pattern 54 near the groove.

As shown in FIG. 8, the upper left of the teaching pattern is set to 1 and the numbers are numbered up to the mth in the horizontal direction and the nth in the vertical direction.
(I, j). Similarly, input patterns 51, 52, 5
The upper left of 3 and 54 are set to 1 and numbered up to the mth in the horizontal direction and up to the nth in the vertical direction, and the luminances at the i-th and j-th positions in the horizontal direction are represented by U1 (i, j) and U2 (i,
j), U3 (i, j) and U4 (i, j). Correlation coefficient R between teaching pattern luminance W and input pattern luminance U
To

[0015]

(Equation 6)

Is calculated.

The correlation coefficients R1, R2, R3 and R4 between the teaching pattern and the input patterns 51 to 54 are R4> R1, R4> R2, R4> R3, and the input pattern 54 near the groove (FIG. 1) , FIG. 7) has the largest correlation coefficient, and the groove can be determined well.

This corresponds to the teaching pattern 50 (FIGS. 2 and 3).
3 (b) in FIG. 3 and a1, b1 of the input pattern 51 (FIGS. 1, 4) near the sputter.
4 (b) of FIG. 4, the luminance waveform between a2 and b2 of the input pattern 52 (FIGS. 1 and 5) near the paint character (FIG. 5 (b)), and the input near the tack welding. Pattern 5
3 (FIGS. 1 and 6) and a luminance waveform between a4 and b4 of the input pattern 54 (FIGS. 1 and 7) near the groove (FIG. 7B). (B)), it can be seen that the input pattern 54 is most similar to the teaching pattern 50 and has a high correlation coefficient.

Next, in the one-dimensional groove position detection method described in (2) (corresponding to claim 2), the groove is photographed in the vertical direction before the scanning control and stored in the memory (s × t pixels). The gray image data of the section consisting of the groove and the surface of the weld plate is extracted from the gray image (FIG. 2) obtained (FIG. 9). The upper left of the extraction area (FIG. 9) is set to 1 and numbered up to t in the horizontal direction and up to s in the vertical direction, and the luminance at the x-th and y-th positions in the horizontal direction is represented by B.
When expressed as (x, y), the y-th luminance distribution C (x) in the horizontal direction (x direction) is expressed as C (x) = B (x, y) (7). The luminance distribution C (x) in the horizontal direction (x direction) when the luminance is averaged in the vertical direction (y direction) is

[0020]

(Equation 8)

## EQU1 ## Luminance distribution C from x = 1 to t
FIG. 10 shows the waveform of (x). P is selected such that the groove portion with low luminance is at the center, and p is selected from the p-th position from the left end in FIG.
The luminance waveform data up to the + m-th luminance pattern is a teaching pattern D (i) D (i) = C (p + i) (9) where i = 1 to m. FIG. 11 shows the teaching pattern D (i).

Similarly, the sputter 2 is included in the image of the vertical groove 21 and the surroundings (FIG. 1) taken during the copying control.
3, the upper left of the image (FIG. 1) composed of s × t pixels in which the paint character 24 and the tack weld 25 are photographed
Then, numbers are assigned up to the tth in the horizontal direction and up to the sth in the vertical direction, and those images are extracted. If the luminance at the x-th position in the horizontal direction and the y-th position in the vertical direction in these images is B (x, y) (FIG. 9), the luminance distribution C in the y-direction in the vertical direction is obtained.
(X) is given by equation (7). The luminance distribution C (x) in the x direction when the luminance is averaged in the vertical direction (y direction) is (8)
It becomes an expression. FIG. 12 shows the waveform of the luminance distribution C (x) from x = 1 to t.

[0023] q + m from any qth from the left end of FIG.
Up to the luminance waveform data input pattern (teaching pattern D
(I), the correlation coefficient between the teaching pattern and the input pattern is

[0024]

(Equation 10)

From FIG. 12, x = q to q + m shown in FIG.
The input pattern up to has the largest correlation coefficient. This can be understood from a comparison between the waveforms of FIGS. 11 and 12 because the waveforms from x = q to q + m in FIG. 12 are most similar to the waveforms in FIG.

As described above, if a groove is photographed on the screen, the maximum value of the correlation coefficient is high, and the groove can be detected by selecting this input pattern. However, when the groove is not photographed on the screen, such as when the tack welding is long, the maximum value of the correlation coefficient is low, and even if this input pattern is selected, the groove is not obtained. Therefore, a reference value for the correlation coefficient is provided, and when the maximum value of the correlation coefficient is lower than the reference value, it is determined that there is no groove.

The following basic experiment was performed to determine the reference value of the correlation coefficient. Temporary welding 2 about 30 mm long and about 100 mm apart on the groove of I groove as shown in FIG.
The groove was continuously photographed from the start to the end in a 25 mm × 25 mm field of view with a television camera directly above the 1000 mm long test plate subjected to Step 5, and recorded with a video tape recorder. The beveled portion 10 × 10 mm of the recorded start screen was used as a teaching pattern. 39 screens (25
Table 1 shows the maximum value of the correlation coefficient of the input pattern of 10 × 10 mm and the position of the input pattern for each screen of “25 × field of view not overlapping”.

[0028]

[Table 1]

From Table 1, when the correlation coefficient is 0.6 or more, the groove position is detected, and when the correlation coefficient is less than 0.6, it mainly indicates the position of the tack bead and the sputter scar. Therefore, the reference value of the correlation coefficient is 0.
6 and 0.6 or more were defined as “grooves” and less than 0.6 as “no groove”.

Next, the change in the correlation coefficient when the width of the gap being controlled is different from the width of the gap taught ((a) in FIG. 3 and FIG. 13) was examined. As shown in FIG. 18, a test plate having the above-mentioned size and having no tack weld 25, and a television camera having a visual field of 25 mm × 25 mm immediately above an arbitrary position of a groove having a surface gap width of 0.2 mm (almost adhered). The groove was photographed, and the groove portion of the screen, 10 × 10 mm (so that the groove was at the center) was used as the first teaching pattern. Similarly, on a test plate of the size shown in FIG. 18 without tack welding, a groove was photographed with a television camera in a visual field of 25 mm × 25 mm immediately above an arbitrary position of a groove having a surface gap width of 2.7 mm. The groove portion 10 × 10 mm (with the groove at the center) was used as the second teaching pattern.

A test plate having the size shown in FIG. 18 and having no tack weld 25, having a surface gap width of 0.2 mm, 0.7 mm, 1.2 mm,
1.7mm, 2.2mm and 2.7mm test plates were prepared.
A groove was photographed with a television camera in a visual field of 25 mm × 25 mm immediately above an arbitrary position on each type of test plate, and a 10 × 10 mm visual field at an arbitrary position on the screen was used as an input pattern.

The maximum value of the correlation coefficient of the first teaching pattern is determined while moving the position of the input pattern in each screen, and the maximum value of the correlation coefficient corresponding to each gap width is calculated.
It is shown in Table 2. Similarly, the maximum value of the correlation coefficient of the second teaching pattern was obtained while moving the position of the input pattern in each screen, and the maximum value of the correlation coefficient corresponding to each gap width was obtained. Table 2 shows the results.

[0033]

[Table 2]

As can be seen from Table 2, when the gap width of the input pattern changes with respect to the gap width of the teaching pattern, the correlation coefficient decreases. The correlation coefficient became less than 0.6 when the gap width increased or decreased more than 2.0 mm.

From the test results of the groove where the tack welding 25 was performed, the correlation coefficient must not be less than 0.6 with respect to the change in the gap. Therefore, in the method using a plurality of teaching patterns having different gap widths of the teaching pattern, in the case of the groove shown in Table 2, the gap is 0.2 mm.
If a correlation coefficient is calculated for two teaching patterns of 2.7 mm and the higher correlation coefficient is adopted, the groove position can be detected more accurately with a correlation coefficient of 0.7 to 1.0.

In the case where the amount of displacement of the input pattern being controlled is different from the amount of displacement of the teaching pattern (see FIG. 1).
The change in the correlation coefficient of 4) was investigated. In the test plate of the above-mentioned size shown in FIG. 18, a groove having a surface gap width of 0.2 mm was linearly traced with a television camera from just above an arbitrary position on a 25 mm × 25 mm field of view without any displacement. An image was taken, and a groove portion 10 × 10 mm (so that the groove was at the center) of the screen was used as a first teaching pattern. Similarly, the surface gap width was 0.2 mm with a test plate having the above-mentioned size and having no tack weld shown in FIG.
A groove is taken by a TV camera from a position directly above an arbitrary position on the test plate with a visual field of 25 mm × 25 mm at a shift amount of 160/1000, and the groove portion of the screen is 10 × 10 mm (groove is (To be in the center) as the second teaching pattern.

In the test plate having the above-mentioned size shown in FIG. 18, a groove having a surface gap width of 0.2 mm was moved along a linear scanning deviation amount of 0, 40/1000, and
80/1000, 120/1000 and 160/1000, 25mm × 2 from just above any position on any of the five types of test plates with a television camera
Shoot a groove with a 5mm field of view and 10 ×
A 10 mm field of view was used as the input pattern. Table 3 shows the result of finding the maximum value of the correlation coefficient of the first teaching pattern while moving the position of the input pattern in each screen, and finding the maximum value of the correlation coefficient according to each line scanning deviation amount. Similarly, the maximum value of the correlation coefficient of the second teaching pattern was obtained while moving the position of the input pattern in each screen, and the maximum value of the correlation coefficient corresponding to each line scanning deviation amount was obtained. Table 3 shows the results.

[0038]

[Table 3]

From Table 3, it can be seen that the correlation coefficient is reduced when the line scanning deviation of the input pattern is changed with respect to the line deviation of the teaching pattern. The maximum value of the correlation coefficient of the first teaching pattern (the linear scanning shift amount = 0) was 0.6 or more up to the input pattern of 120/1000, but was less than 0.6 at 160/1000. The maximum value of the correlation coefficient was 0.6 or more when the input pattern was 40/1000 or more in the second teaching pattern (line copying shift amount = 160/1000), but was less than 0.6 when the input pattern was 0.

From the test results of the groove where the tack welding was performed, the correlation coefficient must not be less than 0.6. Therefore, in the method using a plurality of teaching patterns having different line copying displacements, in the case of the groove shown in Table 3, the line copying displacement of the first teaching pattern = 0 and the line copying displacement of the second teaching pattern = 160. / 1000, and further, taking into account the direction of the line scanning deviation, calculate the correlation coefficient with the input pattern for the three teaching patterns of the line scanning deviation amount = -160/1000 as the third teaching pattern, and determine the higher correlation coefficient. adopt. The 0 and 40/1000 line scanning deviations have a high correlation coefficient with the first teaching pattern.
0.8 or more, the line copying deviation amounts 120/1000 and 160/1000 have a high correlation coefficient of the second teaching pattern.
The correlation coefficient between the teaching pattern and the second teaching pattern is equal to 0.7 to 0.8, and the groove position can be more accurately detected with a large correlation coefficient of 0.7 to 1.0 as a whole for the linear scanning deviation amount 0 to 160/1000.

As described above, the method of determining the pattern coincidence from the correlation coefficient between the teaching pattern and the input pattern is as follows.
There is an essential problem that the correlation coefficient decreases with changes in the size and inclination of the target image, and erroneous determination. In order to avoid this problem, a correlation coefficient is calculated for a plurality of teaching patterns prepared in advance in consideration of a variation range of a size (a width of a gap of a groove) and an inclination (amount of a line scanning deviation) of a target image, The judgment method is very effective.

The groove that is shifted according to the line is drawn obliquely on the screen (FIG. 15). If a groove with a line scanning deviation of 160/1000 is photographed with a visual field of 25 × 25 mm, the groove position at the upper end and the lower end of the screen is different by about 4 mm. Therefore, when the image is captured as shown in FIG. 15, it is necessary to correct the position of the welding torch in the left-right direction according to the position of the detected groove position in the vertical direction on the screen. In the case of performing welding while controlling the groove following, it is necessary to obtain a groove position at a welding torch (arc generation point). Usually, since the front of the welding torch 3 is photographed, the input pattern 53 near the groove position detected on the screen is used.
Is an input pattern near the groove position at the control point P (welding torch 3). That is, as shown in FIG. 15, a gray scale screen (18 in FIG. 15) in the vicinity of a groove during contour control, which is composed of horizontal t and vertical (traveling direction) s pixels, is set, and a reference point is set at an arbitrary position on the same plane. O is provided with a t axis parallel to the horizontal direction of the screen 18 and an s axis parallel to the vertical direction. The upper left coordinate of the screen 18 is set to (t0, s0) in correspondence with the horizontal pixel pitch and the vertical pixel pitch of the screen 18, The upper right coordinate is (t0 + t, s
0), the lower left coordinate is (t0, s0 + s), and the lower right coordinate is (t
0 + t, s0 + s). The coordinates of the control point P (welding torch 3) are set to (t2, s2). A method of correcting the groove detected on the screen 18 so as to be the groove position at the control point P is as follows. If the coordinates of the detected groove position are (t1, s1), then the line trace shift amount T of the teaching pattern is obtained. Is known in advance, the coordinates (t3, s2) of the groove position after correction at the control point P must be controlled so that t3 = t1− (s2−s1) × T (11) No.

Further, a teaching pattern was used in the case where the two plates constituting the groove had different levels of light and shade (paints having different thicknesses were painted, or the storage conditions after painting were different, etc.). The effect of the correlation coefficient was investigated. As described above, since the image signal voltage changes in accordance with the density of the surface of the plate as described above, the digital value after analog / digital conversion of the image signal voltage is large on the light plate surface and small within the deep groove. This digital value differs depending on the number of bits at the time of analog / digital conversion.
The brightest (white) digital value is 255, and the darkest (black) digital value is 0, which is a value between 0 and 255. Prior to the investigation, steel plates having different periods of time after paint painting and different coating film thicknesses were prepared, photographed while illuminating the surface of the steel plate with a certain brightness, and analog / digital converted the image signal. At this time, the digital value of steel sheet A was 11 on average.
0, steel plate B had an average of 150, and steel plate C had an average of 160.

First, a steel plate A was used for both of the two steel plates, and a test plate having the above-mentioned size shown in FIG.
Create an I-groove of 0 mm, and shoot the groove with a TV camera in a visual field of 25 mm x 25 mm from immediately above any position on the groove,
10 × 10mm groove on the screen (so that the groove is in the center)
Is the first teaching pattern. Similarly, using the steel plates A and B and the test plate having the above-described size shown in FIG.
Create an I-groove of 1.0mm, shoot the groove with a 25mm x 25mm visual field with a TV camera from just above any position of the groove, and make a groove of 10 x 10mm (the groove is the center) on the screen ) As the second teaching pattern. Further, using steel plates A and C, an I groove having a surface gap width of 1.0 mm was created with a test plate having the above-described size shown in FIG. 18 and a 25 mm × 25 mm TV camera was formed immediately above an arbitrary position of the groove. The groove was photographed in the visual field, and the groove portion of the screen, 10 × 10 mm (so that the groove was at the center) was used as the third teaching pattern. A steel plate C was used for both of the steel plates, an I groove having a surface gap width of 1.0 mm was created with a test plate having the size shown in FIG. 18, and a field of view of 25 mm × 25 mm was observed with a television camera immediately above any position of the groove. The groove was photographed by using, and the groove portion of the screen, 10 × 10 mm (so that the groove was at the center) was used as the fourth teaching pattern.

Next, the steel sheet A was used for both of the two steel sheets.
A groove having a surface gap width of 1.0 mm was created using a test plate having the size shown in FIG. 18, and a groove was photographed with a television camera in a visual field of 25 mm × 25 mm immediately above an arbitrary position of the groove. A 10 × 10 mm field of view at the position was used as the input pattern.
The result obtained by calculating the maximum value of the correlation coefficient of the first teaching pattern while moving the position of the input pattern in the screen is shown in Experiment 4 of Table 4. It is shown in FIG. Here, D1 and D2 are the average values of the digital values after analog / digital conversion of the two plate surfaces forming the groove of the teaching pattern. However, analog / digital conversion is performed in 8 bits, and the brightest (white)
The digital value is 255, the darkest (black) digital value is 0
The grayscale image was subjected to analog / digital conversion so as to have a value between 0 and 255.

Similarly, Table 4 shows the results of finding the maximum values of the correlation coefficients of the second teaching pattern, the third teaching pattern, and the fourth teaching pattern. 2, No. 3 and No. Shown as 4. Further, the steel sheet C is used for both of the two steel sheets, and FIG.
Create a groove with a surface gap width of 1.0 mm using a test plate of the size shown in
The groove was photographed in a visual field of mm × 25 mm, and a visual field of 10 × 10 mm at an arbitrary position on the screen was used as an input pattern. The result obtained by calculating the maximum value of the correlation coefficient of the first teaching pattern while moving the position of the input pattern in the screen is shown in Experiment 4 of Table 4. It is shown in FIG. Similarly, Table 4 shows the result of finding the maximum value of the correlation coefficient of the second teaching pattern, the third teaching pattern, and the fourth teaching pattern. 2, No. 3 and No. Shown as 4.

[0047]

[Table 4]

In Table 4, No. From the data of No. 1, the teaching pattern is steel sheet A for both of the two steel sheets, and the difference in shading of the steel sheet surface is 0.
In the case of, the correlation coefficient is high if the difference in shading of the steel sheet surface is 0 regardless of whether the input pattern is steel sheet A for both of the two steel sheets or steel sheet C for both of the two steel sheets. No. of Table 4 From 4, the teaching pattern is the case where both the two steel sheets are steel sheet C and the difference in shading of the steel sheet surface is 0, and the input pattern is that both the steel sheet A and the two steel sheets are both steel sheet C regardless of the input pattern. If the difference in shading is 0, the correlation coefficient is high. However, in Table 4, No. 2 and No. As can be seen from FIG. 3, the correlation coefficient was lower in the teaching pattern in which the difference in shading between the two steel plates constituting the groove was 0.16 and 0.20, and in the input pattern in which the shading difference between the two steel plates was 0. Therefore, when two steel plates are equally light or dark, the correlation coefficient does not decrease, but there is a characteristic that the correlation coefficient decreases when the difference between the light and dark changes.

In actual welding, a groove formed of a steel plate of several meters to several tens of meters may cause variations in the coating film thickness, differences in storage conditions, and cases where the plate is joined in the middle of the length direction. ,
The lightness and darkness of the steel sheet surface differ from those of the two steel sheets. But,
It is unpredictable which will be darker or lighter. Therefore, the teaching pattern is preferably a groove having a small difference in shading on the surface of the two steel sheets, but a good teaching pattern may not always be obtained with an actual groove due to variations in shading. Therefore, we prepared a pseudo groove with a dark line corresponding to the groove on a light background corresponding to the density of the two steel sheets,
By taking this as a teaching pattern instead of taking a groove, stable copying control can be performed. Alternatively, by using a method in which a digital value corresponding to the density of the welding plate surface and the groove is directly written into the memory address of the teaching pattern corresponding to the position of the welding plate surface and the groove and used as a teaching pattern, stable copying control can be performed. Will be possible.

Further, the difference in density between the surfaces of the two steel plates forming the groove greatly changes, 0.16 <(D1-D2).
/ 255 or (D1−D2) / 255 <−0.16 (12) In order to keep the correlation coefficient from being less than 0.6, the density of the two steel sheets is determined as a teaching pattern. Preparing the one satisfying the expression (13) with a small difference, -0.16 ≦ (D1−D2) /255≦0.16 (13) Assuming a large difference from the teaching pattern (14) Prepare one that satisfies the formula, and set 0.16 <(D1-D2) / 255.
And D1-D2 / 255 <-0.16 (14) Calculate a correlation coefficient for a plurality of teaching patterns, and detect a groove in a range where the correlation coefficient does not decrease if a higher correlation coefficient is adopted. Can be.

The number of pixels is desirably 3 to 100 pixels / mm. The copying accuracy of arc welding must be within ± 1 mm, and if it is less than 3 pixels / mm, the resolution is low and the copying accuracy cannot be obtained. One screen of a general television camera has 525 scanning lines.
Therefore, the number of pixels in one screen in the vertical direction is 525 at the maximum.
At 3 pixels / mm, the vertical field of view is 17
5 mm. If the number of pixels exceeds 100 pixels / mm, the field of view becomes narrower to 5.25 mm. In the teaching pattern and the input pattern, both the groove and the surface of the welding plate need to be included in the pattern. Gas cutting, plasma cutting,
The surface gap of the groove is widened depending on the cutting method or cutting conditions such as laser cutting, so that both the groove and the surface of the weld plate are prevented from entering the visual field.

The scanning line of one screen of a general television camera is 5
Since the aspect ratio of the screen is 3: 4 with 25 lines, the total number of pixels is preferably 525 × 700 or less. From the characteristics of the computer, 2 × n powers of 256 × 256 or 512 × 5
In many cases, the total number of pixels is 12.

[0053]

FIG. 16 shows the configuration of an automatic welding apparatus used in an embodiment of the present invention, and FIG. 17 shows the configuration of a control apparatus. A traveling carriage 1 equipped with a scanning shaft 4 to which a television camera 2 and a welding torch 3 are attached, and an image signal from the television camera 2 are taken in and converted by an A / D converter 11 to produce one screen of image data of 256 screens. Pixel x 256 horizontal pixels, total 6553
An image memory 12 for storing image data for six pixels, a drive motor 13 for processing the image data and for driving the scanning shaft 4
And a microcomputer 15 for instructing a servo amplifier 14 for controlling the control. The data in the image memory 12 was observed on a monitor television 17 via a D / A converter 16 as necessary.

Example 1 A groove 21 was photographed at the center of the screen 18 (FIG. 17) and stored in the image memory 12. Image data of 50 pixels vertically by 50 pixels horizontally in the center of the screen 18 was stored as a teaching pattern in the microcomputer 15. The groove is I groove as shown in FIG.
Approx.
Welding was started by running the traveling trolley 1 having a groove formed by shifting the line copying deviation by 10 mm with respect to 00 mm.

The groove image photographed in accordance with the command of the microcomputer 15 is stored in the image memory 12. The microcomputer 15 sequentially takes out the input pattern of 50 × 50 pixels from the image memory 12, The input pattern with the largest correlation coefficient less than 0.6, which is calculated according to the equation (6), is determined to have no groove, and the input pattern is not referred to in the copying control. An input pattern having a relation number of 0.6 or more was determined to be a groove, and the copying was controlled in such a manner that the groove 21 of the input pattern was in the center of the screen 18.

Embodiment 2 The groove 21 is photographed at the center of the screen 18 and the image memory 1 is taken.
2 memorized. The image data was added in the vertical direction of the screen 18 and stored in the microcomputer 15 as a luminance waveform in the groove width direction (x direction). Further, a luminance waveform of 50 pixels at the center of the screen 18 in the groove width direction (x direction) is stored in the microcomputer 15 as a groove teaching pattern. The groove is an I groove as shown in FIG. About 30m above
Approximately 100mm intervals, tack welding 25 at intervals of about 100mm, length 100
A groove was created by making the line copying deviation 10 mm with respect to 0 mm.
The traveling trolley 1 was run to start welding.

The groove image photographed in response to the instruction of the microcomputer 15 was stored in the image memory 12. The image data is added in the vertical direction of the screen according to equation (8) and stored in the microcomputer 15 as a brightness waveform in the groove width direction (x direction). The microcomputer 15 stores the left end of the stored brightness waveform in the x direction. , The waveform data for 50 pixels is taken out as an input pattern, and the waveform data for 50 pixels is taken out as an input pattern by sequentially shifting it to the right by one pixel, and the correlation coefficient with the teaching pattern is calculated according to equation (10). An input pattern with a large correlation coefficient of less than 0.6 is determined as having no groove, and is not used for scanning control. An input pattern with a maximum correlation coefficient of 0.6 or more is determined as a groove, and the input pattern is determined. The contour control was performed in such a direction that the groove 21 of the screen became the center of the screen 18.

-Example 3-Setting of a teaching pattern: An I groove with a gap width of 0.2 mm is prepared on the test plate shown in FIG. 18, and the groove 21 is perpendicular to the center of the screen 18 (line scanning displacement amount = 0). And store it in the image memory 12 at the center of the screen 12.
Image data of 50 horizontal pixels was stored in the microcomputer 15 as a first teaching pattern. Furthermore, groove 21
Is captured at the center of the screen 18 with a line scanning deviation amount of 160/1000 and stored in the image memory 12, and the image data of 50 × 50 pixels is stored in the microcomputer 15 at the center of the screen 12 in the microcomputer 15. It was memorized as. Further, the groove 21 is shifted by -16 at the center of the screen 18.
The image was taken as 0/1000 and stored in the image memory 12. Image data of 50 pixels vertically × 50 pixels horizontally was stored in the microcomputer 15 as the third teaching pattern in the center of the screen 12.

Similarly, the test plate shown in FIG.
A 2.7 mm I groove is prepared, and the groove 21 is vertically photographed at the center of the screen 18 (the amount of line shift is 0) and stored in the image memory 12.
The image data of 0 pixel was stored in the microcomputer 15 as a fourth teaching pattern. Further, the groove 21 is photographed at the center of the screen 18 with a line copying deviation amount of 160/1000 and stored in the image memory 12, and the image data of 50 × 50 pixels is stored in the center of the screen 12 in the microcomputer 15. Is stored as the fifth teaching pattern. Further, the groove 21 is shifted by -160 /
Shoot as 1000 and store it in the image memory 12,
At the center of the screen 12, image data of 50 pixels vertically by 50 pixels horizontally was stored in the microcomputer 15 as a sixth teaching pattern.

In the test plate shown in FIG. 18, the gap width was continuously changed from 0 to 3 mm, and an I groove prepared by tack welding at a length of about 30 mm and a pitch of 100 mm was prepared.
The test plate was set so as to be 160/1000, the traveling trolley 1 was run, and welding was started. The groove image photographed in response to the instruction of the microcomputer 15 was stored in the image memory 12. An input pattern of 50 pixels vertically × 50 pixels horizontally is sequentially taken out from the image memory 12, and the correlation coefficient with the teaching pattern is calculated according to the equation (6). The largest correlation coefficient value and the pattern position on the screen are stored. Are repeated from the first teaching pattern to the sixth teaching pattern.

An input pattern in which the largest value of the correlation coefficient from the first teaching pattern to the sixth teaching pattern is less than 0.6 is determined as having no groove, and is not used for copying control. The input pattern with the largest value of 0.6 or more is determined to be a groove,
The correction was performed according to the equation (11), and the copying was controlled in such a manner that the center of the groove 21 of the screen 18 in the length direction became the center of the width of the screen 18.

-Example 4-Setting of a teaching pattern: A teaching groove was created by directly writing a digital value after analog / digital conversion into a memory of a teaching pattern as a pseudo groove. 140 was stored in a memory address corresponding to the surface of the two steel sheets as a digital value corresponding to the shading of the surface of the two steel sheets. Similarly, as a digital value corresponding to the shading in the groove, 40 was stored in a memory address corresponding to the groove. The gap width of the groove was set to 10 pixels so as to correspond to an actual 1 mm.

Using a steel plate A and a steel plate B, an I groove prepared by temporary welding at a gap width of 1 mm, a length of about 30 mm, and a pitch of 100 mm using the test plate shown in FIG. 18 was prepared.
The test plate was set so as to be 10/1000, the traveling trolley 1 was run, welding was started, and a groove image photographed in response to a command from the microcomputer 15 was stored in the image memory 12. An input pattern of 50 pixels vertically × 50 pixels horizontally is sequentially extracted from the image memory 12, and the correlation coefficient with the teaching pattern is calculated according to the equation (6). The input pattern having the largest correlation coefficient less than 0.6 is a groove. An input pattern having the largest correlation coefficient of 0.6 or more was determined to be a groove, and scanning control was performed in a direction in which the groove 21 was located at the center of the screen 18.

-Comparative Example 1- This is a method based on binarization of a conventional method, in which a groove 21 is photographed in the center of a screen 18 and an image signal is binarized via a comparator 19, and the image memory 1
2 memorized. The threshold value of the comparator 19 was adjusted so that the groove became black and the periphery became white. The groove was I groove as shown in FIG.
Approximately 100mm intervals, tack welding 25 at intervals of about 100mm, length 100
Welding was started by running the traveling trolley 1 in which a groove was created by changing the line deviation by 10 mm from 0 mm.

The groove image photographed in accordance with the command of the microcomputer 15 is binarized via the comparator 19 and stored in the image memory 12. The microcomputer 15 corresponds to the groove black from the left end of the screen. Image data is read out from the image memory 12 at the center up to the pixel, and the number of pixels NL from the left end is counted. Similarly, image data is read out from the image memory 12 at the center up to the pixel corresponding to the groove black from the right end. When the number of pixels NR is counted, if NL + NR> 256, it is determined that there is no groove, and copying control is not performed. If NL + NR <256, it is determined that there is a groove, and NL
Copying control was performed in a direction where −NR = 0.

Comparative Example 2 In this comparative example, the same groove as that of the embodiment is controlled under the same conditions by using only the first teaching pattern of the third embodiment.

In the above Examples and Comparative Examples, welding was performed by latent arc welding and carbon dioxide gas shielded arc welding. For the latent arc welding, a welding torch 5 was used for the latent arc welding capable of spreading the flux, and for the carbon dioxide gas shielded arc welding, a welding torch for the gas shielded arc welding was used. As shown in FIG. 18, the groove was an I-groove, which was tack-welded 25 on the groove at an interval of about 30 mm and at an interval of about 100 mm. The evaluation method is
A constant voltage power supply and a pen recorder were connected to a potentiometer 5 attached to the scanning shaft 4, and a scanning control process during welding was recorded. At this time, it was evaluated that the maximum copying deviation was within ± 1 mm.

Table 5 shows the results of the operation. The method of the present invention relates to Examples 1, 2, 3, and 4.
Both latent arc welding and carbon dioxide gas shielded arc welding were good. However, in Comparative Example 1 using binarization, the groove becomes unclear due to a change in the brightness of the welded plate during welding, a scorch mark such as spatter during tack welding, and the recognition of the groove position is incorrect. I was In carbon dioxide gas shielded arc welding, the influence of the arc light was great, and good groove position measurement was not possible. In Comparative Example 2, the maximum value of the correlation coefficient was reduced due to the fluctuation of the gap width of the groove or the line deviation, and the position showing the maximum value was an incorrect groove position such as a scorch mark or a scratch due to spatters other than the groove. Detection was being performed.

[0069]

[Table 5]

[0070]

According to the present invention, since the groove position detection based on the image can be detected satisfactorily by a method completely different from the conventional method, the reliability of the automatic profiling control using the television camera is greatly improved, and the welding process Can be automated and unmanned.

[Brief description of the drawings]

FIG. 1 is a plan view showing an example of a screen obtained by photographing a groove with a television camera.

FIG. 2 is a plan view showing a screen obtained by photographing a groove portion having no disturbance with a television camera.

FIG. 3A is a teaching pattern cut out from a shooting screen;
(B) is a plan view of the teaching pattern.
5 is a graph showing a luminance distribution between the two.

FIG. 4A is a plan view showing an input pattern near a sputter cut out from a photographing screen, and FIG. 4B is a view showing the input pattern.
6 is a graph showing a luminance distribution between a1 and b1 on the screen.

5A is a plan view showing an input pattern near a paint character cut out from a photographing screen, and FIG. 5B is a graph showing a luminance distribution between a2 and b2 on the input pattern.

FIG. 6A is a plan view showing an input pattern in the vicinity of tack welding cut out from a photographing screen, and FIG. 6B is a graph showing a luminance distribution between the input patterns a3 and b3.

7A is a plan view showing an input pattern near a groove cut out from a photographing screen, and FIG. 7B is a graph showing a luminance distribution between a4 and b4 of the input pattern.

FIG. 8 is a plan view showing an image area (size) common to a teaching pattern and an input pattern.

FIG. 9 is a plan view showing a processing target image area (size) set on a screen shot during the copying control.

FIG. 10 is a graph showing a luminance waveform C (x) obtained by adding grayscale image data in the vicinity of a groove taken before the copying control in the vertical direction.

FIG. 11 is a graph showing a teaching pattern D (i) cut out from the graph shown in FIG. 10;

FIG. 12 is a graph showing a luminance waveform C (x) obtained by vertically adding image data captured during the copying control.

FIG. 13 is a plan view showing an input pattern in which a groove gap width is wide.

FIG. 14 is a plan view showing an input pattern shifted in line scanning.

FIG. 15 is a plan view showing a groove image shifted in line scanning.

FIG. 16 is a perspective view showing an external appearance of a mechanism section of an apparatus for implementing the present invention in one embodiment.

FIG. 17 is a block diagram illustrating a scanning control system coupled to the scanning welding apparatus illustrated in FIG. 16;

FIG. 18 is a plan view showing two steel plates used in the experiment.

[Explanation of symbols]

1: Traveling trolley 2: TV camera 3: Welding torch 4: Copying shaft 5: Potential 11: A / D converter 12: Image memory 13: Copying shaft motor 14: Servo amplifier 15: Microcomputer 16: D / A converter 17: Monitor TV 18: Screen 19: Comparator 21: Groove 22: Weld plate surface 23: Spatter 24: Paint character 25: Temporary welding 26: Rust 50: Teaching pattern 51: Input pattern near spatter photographed during scanning control 52: Input pattern near the paint character photographed during the copying control 53: Input pattern near the tack weld photographed during the copying control 54: Input pattern near the groove captured during the copying control

────────────────────────────────────────────────── ─── Continuation of front page (56) References JP-A-1-201784 (JP, A) JP-A-62-199270 (JP, A) JP-A-59-209482 (JP, A) JP-A-64 27775 (JP, A) JP-A-55-87283 (JP, A) JP-A-62-261903 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) B23K 9/127 G01B 11 / 00 G01B 11/24

Claims (7)

(57) [Claims] 1. A groove position detecting method for detecting the position of a welding groove using an image signal voltage of a television camera, comprising:
Digitally converting and storing grayscale image data of s × t pixels in the vicinity of the groove before the contour control, which is composed of a two-dimensional image signal voltage, and from the grayscale image data near the groove before the contour control, the groove and the surface of the welding plate. N × m pixels, s> n, t>
m, of the gray-scale image data of the area is stored as teaching pattern, a gray-scale image of the groove near taken during scanning control Analog /
Digitally converting and storing grayscale image data of s × t pixels in the vicinity of the groove during the contour control during the contour control consisting of a two-dimensional image signal voltage, and the same size as the teaching pattern from the gray image data near the groove during the contour control. , N × m pixels, s> n, t
> M, sequentially takes as input the pattern regions of lateral i-th from the top left of the teaching pattern, longitudinally j th
The brightness W (i, j) at the eye position and whether
From the i-th position in the horizontal direction and the j-th position in the vertical direction
j) is calculated by the following equation (6). Wherein a region having a correlation coefficient R of 0.6 to 1.0 and a largest correlation coefficient is set as a groove position. 2. A groove position detecting method for detecting a position of the welding groove by using an image signal voltage of the television camera, the light and shade image of the groove near taken before copying control Analog /
And digital conversion, storing the gray image data, the luminance distribution in the vertical direction y th laterally from the upper left of the concentration pale image data
From the left end consisting of the groove and the surface of the weld plate from C (x)
Storing grayscale image data of the section from the q-th meaning to q + m th as teaching pattern D (i), the light and shade image of the groove near taken during scanning control Analog /
And digital conversion, storing the gray-scale image data, from the upper left of the light and shade image data in the control had該倣longitudinal y th
Lateral direction (x direction) q from any q th from the left end of the teaching pattern equal size from the luminance distribution C (x) of
+ Sequentially takes as input pattern interval m to th, the teaching pattern D (i) and the correlation coefficient R of the image signal voltage of the input pattern below (10) Wherein a region having a correlation coefficient R of 0.6 to 1.0 and a largest correlation coefficient is set as a groove position. 3. A process of storing, as a teaching pattern, gray-scale image data of a region including a groove and a surface of a welding plate from among the gray-scale image data in the vicinity of a groove before profiling control, for each groove having a different gap. The process of storing one groove teaching pattern each time, calculating the correlation between the image signal voltage of the teaching pattern and the image signal voltage of the input pattern, and finding an area having a correlation coefficient of 0.6 to 1.0 and the largest correlation coefficient is performed. Multiple times as many times as
3. The groove position detecting method according to claim 1, wherein an area having a largest correlation coefficient is set as a groove position. 4. A process of storing, as a teaching pattern, grayscale image data of a region including a groove and a surface of a welding plate from among the grayscale image data near a groove before the profiling control, the amount of line tracing deviation or the amount of line tracing deviation and gap. Done for different bevels,
A line scanning deviation amount or a teaching pattern of a groove having a different line scanning deviation amount and a gap is stored one by one, and a correlation between the teaching pattern and the image signal voltage of the input pattern is calculated. Perform the process of finding the region with the largest correlation coefficient multiple times for the number of teaching patterns,
2. The groove position detecting method according to claim 1, wherein an area having a largest correlation coefficient is set as a groove position. 5. A reference point O is provided in advance at one point on the same plane as a gray screen near a groove during contour control, which is composed of t × s pixels of s pixels vertically and t pixels horizontally, which is a running direction. The coordinate position of the groove position detected when the pixel is converted from O is (t1, s1), and the coordinate position of the control point P before correction is (t1, s1).
2, s2), assuming that the line scanning deviation amount of the teaching pattern having the largest correlation coefficient is T, the coordinate position (t3, s2) of the groove position at the control point P after correction is t3 = t1- (s2- 5. The groove position detection method according to claim 4, wherein s1) × T. 6. An average value of digital values after analog / digital conversion of a grayscale image of the surface of two plates constituting a groove of a teaching pattern stored before the scanning control is defined as D1 and D2. Assuming that the digital value of the grayscale image after analog / digital conversion is Db, a pseudo groove drawn to satisfy the following equations (1), (2) and (3) is photographed, and the grayscale image is subjected to analog / digital conversion. And stores gray-scale image data before and after the contour control, which is composed of two-dimensional or one-dimensional image signal voltages, and stores, as a teaching pattern, gray-scale image data of a pseudo groove portion from among the gray-scale image data near the groove before the contour control. Or a pseudo groove satisfying the expressions (1), (2) and (3), a pseudo groove satisfying the expressions (2), (3) and (4), or (2), (3) and (5) Multiple drawn to satisfy the expression , And analog / digital conversion of the grayscale image to obtain two-dimensional or one-dimensional images.
And storing a plurality of patterns as shade pattern data of a pseudo groove portion from among the shade image data before the contour control before and after the contour control, which is composed of three-dimensional image signal voltages. Characterized by
Claim 1, Claim 2, Claim 3, Claim 4, or Claim 5
-0.16 ≦ (D1−D2) / (Dmax−Dmin) ≦ 0.16 (1) Dmin <Db <D1 <Dmax (2) Dmin <Db <D2 <Dmax (3) 0.16 ≦ (D1−D2) / (Dmax−Dmin) ≦ 1.00 (4) −1.00 ≦ (D1−D2) / (Dmax−Dmin) ) ≦ −0.16 (5) where Dmax is the maximum value of the digital value that can be obtained by the analog / digital conversion of the image signal. Dmin: is the minimum value of the digital value that can be obtained by the analog / digital conversion of the image signal. 7. If the digital values indicating the grayscale images of the surfaces of the two plates forming the groove are D1 and D2, and the digital values indicating the grayscale images in the groove are Db, (1), (1) Digital values satisfying the expressions 2) and (3) are used as grayscale image data of a pseudo groove portion, and D1 and D2 are stored in memory addresses corresponding to the surfaces of the two plates forming the groove of the teaching pattern, and the groove. Db is written to and stored in a memory address corresponding to a digital value satisfying the expressions (1), (2) and (3), a digital value satisfying the expressions (2), (3) and (4) or ( Digital values satisfying the expressions (2), (3) and (5) are stored in a memory address corresponding to the surface of two plates constituting the groove of each teaching pattern as grayscale image data of a plurality of pseudo groove portions. D1 and D2 are also set in the groove. 6. A groove position detecting method according to claim 1, wherein Db is written and stored at a Molly address; -0.16≤ (D1-D2). /(Dmax-Dmin)≦0.16 (1) Dmin <Db <D1 <Dmax (2) Dmin <Db <D2 <Dmax (3) 0.16 ≦ (D1-D2) ) / (Dmax−Dmin) ≦ 1.00 (4) −1.00 ≦ (D1−D2) / (Dmax−Dmin) ≦ −0.16 (5) where Dmax: image signal Dmin: Minimum digital value possible in analog / digital conversion of image signals.