JPS6243123B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a method for measuring so-called moiré shapes.
The present invention relates to an automatic data processing method for obtaining a cross-sectional view of a measurement object from a moiré contour pattern.

As is well-known, the so-called contour moiré is a process in which a grating with periodic transmission characteristics in one direction is placed in front of the object to be measured, and light is directed from a point (line) light source through the grating onto the surface of the object to be measured. When a shadow of a grating (deformed grating image) is created on the surface of an object, and the deformed grating image is further passed through the grating and imaged from the same height as the light source, interference fringes are generated as a result of spatial frequency modulation of the deformed grating image and the grating. These interference fringes represent the contour pattern of the object to be measured, like contour lines on a map, and are therefore used for three-dimensional shape measurement. As shown in Fig. 1, a typical method for generating a contour moire is to place a grating 2 immediately in front of the measurement object 1 and to generate a moire by capturing a deformed grating image 3 through the same grating 2. This method is generally called the irradiation method. Note that in this figure, S is a point light source, and T is an imaging point.

The measurement resolution of this moiré shape measurement method is explained as follows from FIG. Figure 2 is a side view of the irradiation method shown in Figure 1, and shows the point light source S.
The light irradiated from the surface passes through the grating 2 and is irradiated radially onto the measurement object 1 (radiation solid lines starting from S in Figure 2), and the area on the surface of the measurement object 1 that is hit by this light is bright. Also, areas that are not hit by the light become dark, creating a so-called deformed lattice image. When viewed from the imaging point T, the light from the surface of the measurement object is blocked by the grating 2, and only radial light starting from the imaging point T (radial broken lines starting from T in Figure 2) is visible. As a result, the intersection points of the two radiation groups on the surface of the measurement object 1 are imaged brightly, and the other areas are imaged darkly, and this brightness and darkness becomes the brightness and darkness of the moiré fringes. In the figure, the intersection of two radiation groups, that is, the place where bright-phase moiré fringes are formed, is a straight line parallel to the grating 2 (N=1, 2,
...), so the moiré fringes are contour lines from the grid 2.

In FIG. 2, the distance Z ₁ from the grid 2 of the place where bright-phase moiré fringes occur (N=1, 2,...),
Z ₂ , . . . are expressed by equation (1) based on the distance d between the point light source S and the imaging point T, the length l of the point light source S and the imaging point T from the grating 2, and the pitch p of the grating 2.

From equation (1), the resolution of the moire measurement method, that is, the fringe interval ΔZ _N (=Z _N −Z _N-1 ) of the moire fringes, is determined, and the distances of the moire fringes from the grating 2 are also quantitatively calculated. Ru.

By the way, in the conventional moire measurement method, its application was mainly limited to off-line applications, so the measurement results were recorded as a photograph or TV image as a moire contour pattern, and the recorded moire fringe pattern From there, humans extracted 3D information about the measurement target as needed. This is the same procedure as when we look at contour lines on a map and recognize where the mountains and valleys are (qualitative information) or how big the mountains and valleys are (quantitative information). However, such data processing makes it impossible to use moiré shape measurement technology online. Online use here refers to automatic online data processing of shape measurement results using moiré contour patterns into information that can be fed back to the manipulated variables for shape control of the measurement target, for example. The reasons why it was not possible to use it online are as follows.

The first reason is that the moiré fringe pattern is just a contour pattern, and when a contour pattern like the one shown in Figure 3 appears, it is impossible to tell whether it is a convex part or a concave part of the measurement target. . In offline applications, humans can determine whether a part is a convex or concave part by touching the object to be measured, or depending on the object to be measured, it may be known in advance whether it is a convex part or a concave part. This is not a problem in these offline cases, but in online cases where the shape of the object to be measured is unknown in advance and measurement results are required in a short period of time, it becomes a problem if irregularities cannot be determined.

The second reason is that when you want to use automatic data processing to numerically extract the unevenness position and amount of unevenness of a measurement object from a moiré fringe pattern, it is difficult to know whether the measurement object is convex or concave. However, it is difficult to calculate the position and amount of unevenness using automatic data processing. Humans recognize the position and amount of unevenness from a moire fringe pattern. For example, in the case of a moire fringe pattern like the one shown in FIG. 3 (assuming it is convex), the center of the innermost closed curve 4a is the vertex. And, the center height is recognized from the number of closed curves 4a to 4d, but all of these processes rely on the excellent pattern recognition ability of humans, and can be replaced with online automatic data processing. is extremely difficult due to the calculation time and amount of calculation. For example, in a moire fringe pattern as shown in FIG. 4A, if information on the cross section of the object to be measured based on the cross section 6 is required, the moire fringe pattern 5 shown in FIG.
Even if you try to extract measurement target cross-section information from the luminance signal (voltage output signal of a television camera, etc.) along the cross section 6 of Although it is easy to see that the center of the dark-phase moire fringes is located at a certain point, the order of those moire fringes (N in equation (1)) is unknown, so it is difficult to know where the apex is, so it is difficult to draw a cross-sectional diagram. do not have. Also, the moiré stripe interval ΔZ _N (=Z _N
−Z _N-1 ) is calculated from equation (1), but in practice it is often used under the condition that PN/d≪1 (2) If (2) holds, then equation (1) , since the fringe interval ΔZ _N is ΔZ _N ≒P・l/d (3) (ΔZ _N does not depend on N), the order N of the moire fringe is unnecessary when drawing a cross-sectional view of the measurement object from the moire fringe pattern. It is often the case. However, even in this case, it is impossible to draw a cross-sectional view unless the position of the vertex is known. In other words, if we take Figure 4 C as an example, should the position of point B be higher than point A by P·l/d or lower by P·l/d with respect to point A? I don't know if it should be point B' or not.

For the above reasons, online automatic data processing for extracting a cross-sectional view of a measurement object from a moiré fringe pattern has conventionally been difficult.

The purpose of the present invention is to improve the conventional moire measurement method and eliminate the above-mentioned drawbacks, thereby calculating the cross-sectional shape of the object to be measured in any direction from the moire contour pattern obtained by the moire measurement method by automatic data processing. The point is to make it possible. In order to achieve this objective, the present invention detects the direction in which the moire contour pattern changes by slightly moving the light source and image sensor necessary for generating and imaging moire fringes in the horizontal direction. ,
Automatically calculates the cross-sectional shape of the measurement object in any direction by determining the unevenness of the measurement object.

That is, in the present invention, a grating is disposed in front of an object to be measured, and a light source directed toward the object is provided behind the grating, and the object is irradiated from the light source so that the surface of the object is illuminated by the grating. In moiré shape measurement, which measures the three-dimensional shape of a measurement target as a two-dimensional contour pattern stripe by generating a shadow and imaging the shadow, the object to be measured in the required direction from the obtained moire contour pattern. In order to obtain a cross-sectional shape, the light source or the image sensor is slightly moved within a plane at the same distance from the grating in a direction perpendicular to the grating system or in a direction having a perpendicular component. The method is characterized in that the unevenness of the object to be measured is determined based on the information as to whether the object changes, and the cross-sectional shape of the object to be measured in the required direction is obtained from the moiré contour pattern. Hereinafter, the present invention will be explained in detail with reference to the drawings.

Figure 5 is an example of application of this measurement method to online bulging measurement of CC (continuous casting) slabs. CC slab bulging refers to the two-dimensional bulge of the CC slab. In Fig. 5, this refers to the fact that the solidified shell 7 of the CC slab swells two-dimensionally between the rolls 9 under the static pressure of the unsolidified part 8, which causes defects such as internal scratches. Therefore, online measurement was awaited. In this example, we will introduce bulging, which is this two-dimensional bulge.
In addition to continuously measuring a two-dimensional contour pattern using the moiré measurement method, the cross-sectional shape of the slab in the width direction and longitudinal direction is calculated at regular intervals through automatic data processing.

In FIG. 5, 10 is a continuous wave argon (Ar) laser, and a laser beam 11 outputted from the laser 10 is expanded into a parallel beam with a large diameter by a beam expander 12, and then by a total reflection mirror 13a. After being bent, the beam is turned into a conical beam by the concave lens 14a, and is irradiated onto the coagulation shell 7 to be measured through the grating 2. At this time, the shadow of the grating 2 (deformed grating image) formed on the coagulation shell 7 is passed through the grating 2 and further through the bandpass filter 15 having wavelength selectivity of the laser beam 11, and then the television camera 16 captures the image of the concave lens 14a. When an image is captured from a location at the same height as the imaginary focal point S (point light source), moiré fringes representing the contour pattern of the coagulation shell 7 are captured by the television camera 16, and the video signal is input to the television monitor 17, and the interface 18 is input. The interface 18 AD converts the luminance signal along the necessary cross section of the video signal and sends it to the computer 19.
At the same time, control is performed to move the total reflection mirror 13a up and down at a constant period. In FIG. 5, when the total reflection mirror 13a moves upward, the laser beam is not reflected by the total reflection mirror 13a, but is reflected by the total reflection mirror 13b, and is also expanded into a conical shape by the concave lens 14b. Here, if the imaginary focus S' of the concave lens 14b is set to the same height as S, the moire point light source will move periodically in the horizontal direction in accordance with the vertical movement of the total reflection mirror 13a. In FIG. 5, the purpose of periodically moving the moiré point light source in the horizontal direction is to determine the unevenness of the object to be measured and calculate the movement of the cross-sectional shape in any direction.
The method will be explained in detail below with reference to the drawings.

FIG. 6 is a diagram of the optical arrangement for moiré measurement in FIG. 5, viewed from the side. In Figure 6, when the position of the point light source moves horizontally from S to S',
The position where bright-phase moiré fringes occur is 20,2
It can be seen that the values change from 1, . . . to 20', 21', . This means that d decreases in equation (1), and if only d decreases in equation (1), Z _N
increases, that is, the position where moiré fringes of the same order are formed moves away from the grating 5. If we pay attention to the movement of the moiré fringes, we can see that Fig. 7A,
As shown in A', in the convex part of the measurement target, as _ZN increases, the moiré fringes 22, 23... expand to 22', 23', etc. around the apex of the convex part, and
Alternatively, new moire fringes are created around the apex of the convex portion. On the other hand, in the concave part of the measurement target, the seventh
As shown in Figures B and B', the moiré fringes 22, 23, . . . shrink as shown in 22', 23', . Therefore, the moving direction of the moiré fringes due to the movement of the light source is opposite between the convex and concave parts of the measurement target.
From this direction, it is possible to determine the concavity and convexity and to know the positions of its apex and bottom points.

For example, in order to obtain a cross-sectional shape in a particular direction from a moiré stripe contour pattern, the following may be performed. FIG. 8a shows a luminance signal along a necessary cross section of a moiré fringe pattern of a measurement object having a convex center, 24 is a signal when the light source position is S, and 24' is a signal when the light source position is S'. This is the signal at the time of . 2
Comparing No. 4 and No. 24', the maximum and minimum peak positions naturally shift due to the movement of the moiré fringes as the light source moves, and the direction of the relative shift includes unevenness information. That is, in FIG. 7a,
Since I' has shifted to the left from I, the cross-sectional shape is upward to the right at I. Since J' has shifted to the right from J, the cross-sectional shape is downward to the right at J, and K and K' are at the same position on the left and right. Therefore, it can be determined that the cross-sectional shape is horizontal at K. Therefore, the fourth
This solves the problem of not being able to draw the cross-sectional shape because the direction of the slope of the maximum and minimum peak positions is unknown, as shown in Figure C.

25 and 25' in Fig. 8b are signals obtained by removing noise from raw signals such as 24 and 24' using a bandpass filter when automatically detecting the peak position. Uneven irradiation (low frequency noise) and harmonic moiré,
Fine unevenness (high frequency noise) caused by small scratches on the surface of CC slabs can be removed.

Figure 8c shows the maximum and minimum peak positions obtained from 25 in Figure 8b, the tilt information obtained from the moving direction of the peak positions 25 and 25', and the fringe interval ΔZ _N calculated from equation (3). This is the cross-sectional shape originally plotted.

The above is a data processing method for obtaining a cross-sectional shape in a particular direction from a moiré stripe contour pattern. This method is shown in a block diagram as shown in FIG. 9. Figure 9 shows the IF (interface) 18 in Figure 5 and the data processing computer 1.
This is a block diagram showing the functions of No. 9. In FIG. 9, the IF 18 A/D converts a portion of the television analog signal from the television camera along the required cross section and transfers it to the data processing computer 19. Here, IF18 is a control signal S or a control signal that moves the light source position for automatic data processing.
S′ is output, and the slab size and required cross-section information are input from the setting device. The TV analog signal is synchronized with the light source position control and is the signal when the light source position is S.
The signals at S' are input alternately. The television analog signals at the two light source positions are alternately A/D converted and each is sent to a data processing computer 1.
9, and after removing noise with a bandpass filter, the peak position is detected. The respective peak positions are compared with each other, and the slope information at the peak position is obtained from the movement direction of the left and right peak positions (unevenness discrimination), and the fringe interval Δ is calculated separately.
Together with the information on Z _N (d, l, P and lift-off required for this calculation are input from the setting device), the required cross-sectional shape is obtained. Once the required cross-sectional shape is obtained,
Shape signals can be input immediately to the shape control process computer.

Next, the configuration of the present invention will be explained. However, basically, the method for generating moire fringes will be explained based on the laser moire method using a laser light source, which is disclosed in a previously filed patent (form measurement method using laser moire), as shown in FIG. In the present invention,
For automatic data processing of moiré contour patterns, the moiré light source or image sensor is slightly moved within the same plane from the grating in a direction perpendicular to the grating system or in a direction having a perpendicular component, and the moving direction of the moire contour pattern is determined. In this case, the light source position may be moved or the image sensor position may be moved. However, it is generally easier to move the light source position than to move the image sensor, and since moving the image sensor changes the composition in which the object to be measured is captured, it is more appropriate to move the light source position.

There are various methods for moving the light source position, such as a method of controlling the optical path by vertically moving a reflecting mirror as shown in FIG. 5, and a method of sliding concave lenses 14a and 14b as shown in FIG. 10. In the method shown in FIG. 10, the light source position also changes in the height direction, but in practice there is no problem because the change in the height direction can be ignored compared to the distance between the grating and the point light source. Therefore, the position of the point light source does not necessarily have to be moved within the horizontal plane in a strict sense. As for the distance by which the light source position or the image sensor is moved, it is sufficient to have a slight movement that can detect the moving direction of each moiré fringe. For example, if P = 1 mm, d = 2 m, and l = 2 m in equation (1), the grating system A distance of several tens of mm in the direction orthogonal to is sufficient. In addition, if the light source position or slight movement of the image sensor is difficult due to spatial or environmental reasons, for example, an optical fiber with a lens placed at the end can be used, since the fiber itself is flexible and very thin. It's convenient and easy to make small movements.

The two moire contour patterns obtained as described above are A/D converted by the IF and input into the computer as shown in Fig. 9. and the input of the moiré pattern shall be synchronized. Strictly speaking, the two moiré contour patterns are imaged at different times, but this poses no problem since the position of the light source or the image sensor can be slightly moved in a very short time.

The computer 19 inputs luminance signals along the necessary cross-sections of the two moire patterns, and compares the moire peak positions of the same order of the respective luminance signals, but generally the raw luminance signal contains signals other than the bright and dark phases of the moire. Since it contains noise components, it is removed by a computer. For example, if a fast Fourier transform is performed to extract only the terms near the spatial frequency components of moiré, and then a fast inverse Fourier transform is performed, noise can be removed in an extremely short time. In addition to digital filtering performed by a computer, these noises can be removed using an analog filter from the analog signal.

The peak position of the brightness signal from which the noise has been removed is detected, and the brightness peak positions of moire fringes of the same order in the two moire patterns are compared. Therefore, the direction of inclination (upward right/downward right/horizontal) at each peak position is determined from the direction of change of the peak position and is stored.

Furthermore, since the luminance peak position along the required cross section of one of the two moiré patterns is already known from the peak position described above, the cross-sectional view can be immediately plotted as long as the direction of inclination is known. Note that the distance in the height direction from the adjacent brightness peak can be easily obtained by calculating the stripe spacing ΔZ _N of the moiré fringes in advance. ΔZ _N is a constant calculated approximately using equation (3), but if you want to calculate it more accurately, you should first calculate the order of the moiré fringes formed on the surface of the measurement object by lift-off (grating and measurement). (distance to the object), so it is sufficient to calculate ΔZ _N around that order.

In the CC bulging hot online measurement shown in FIG. Since the shape can be obtained quantitatively and instantaneously, the amount of CC bulging can be continuously measured, and the amount can be fed-packed to the operation settings of CC withdrawal speed and cooling water amount to reduce CC bulging to within 1 (mm). This has made it possible to control the internal cracking rate, reducing the incidence of internal cracks to less than 10% compared to conventional methods.

As explained above, according to the present invention, it is possible to determine the unevenness of a measurement object in moire shape measurement, and there is an advantage that the shape of the measurement object in any direction can be automatically determined.

[Brief explanation of the drawing]

1 to 4 are explanatory diagrams of contour moiré, FIG. 5 is a block diagram showing an embodiment of the present invention, FIGS. 6 to 8 are explanatory diagrams of its operation, and FIG. 9 is an illustration of the same height moiré. FIG. 10 is a block diagram showing another example of moving the light source position. 1: Measurement object, 2: Grid, 3: Deformed grid image,
S: point light source, T: imaging point, 4a to 4d: moire fringe (dark phase), 5: moire fringe pattern, 6: cross section, 7:
Solidified shell, 8: Unsolidified part, 9: Roll, 10:
Ar laser, 11: Laser beam, 12: Beam expander, 13a, 13b: Total reflection mirror, 14a, 14b: Concave lens, 15: Band pass filter, 16: TV camera, 17: TV monitor, 18: Interface, 19 :
Computer, S, S': Imaginary focus of concave lens (point light source), 20: Position where N-th moire fringes are formed by point light source S (Z _N ), 20': Position where N-th moire fringes are formed due to point light source S (Z _N ), 21: Position where N+1-order moiré fringes are formed due to point light source S (Z _N+1 ), 2
1': Position where N+1-order moire fringes are formed due to point light source S' (Z _N+1 ), 22: N-order moire fringes due to point light source S, 22': N-order moire fringes due to point light source S', 2
3: N+1st-order moiré fringe due to point light source S, 23':
N+1st moiré fringe due to point light source S', 24: TV analog raw signal (light source position S), 24': TV analog raw signal (light source position S'), 25: signal after bandpass filter (light source position S), 25': Signal after bandpass filter (light source position S'), 2
6: Cross-sectional shape.

Claims (1)

[Claims] 1 Place a grid in front of the measurement target and
A light source directed toward the measurement object is provided behind the grating, the light source illuminates the measurement object, a shadow of the grating is generated on the surface of the measurement object, and the three-dimensional shape of the measurement object is determined by imaging the shadow. In the moiré shape measuring device configured to measure as two-dimensional contour pattern stripes, the light source or image sensor is used to obtain the cross-sectional shape of the object to be measured in any direction from the obtained moire contour pattern. means for displacing the grating in a direction perpendicular to the grating system or in a direction having a perpendicular component within a plane at the same distance from the grating; What is claimed is: 1. A shape measuring device using moiré, characterized in that it is provided with an arithmetic processing device for determining