JPH03235007A - Speckle length measuring instrument - Google Patents

Abstract

PURPOSE:To obtain a measuring resolution which exceeds the resolution of a non-dimensional image sensor and measure a length with a high accuracy by changing image signals photoelectrically converted by the one-dimensional image sensor at a period according to the scanning period of a speckle pattern due to the one-dimensional image sensor and binary-coding the image signals for every speckle pattern. CONSTITUTION:An output signal from a one-dimensional image sensor 13 is connected to a sample hold circuit 2 via an amplifier 15 and removes noises peculiar to a CCD. Then, the output signal of the circuit 2 is fed to a binary- coding circuit 4 via a gain control amplifier 3 and an image signal S is binary- coded. Binary-coded data are fed to a correlator 5 wherein the mutual correlation function of a speckle pattern is repeatedly calculated and a result is fed to a microcomputer 7. The microcomputer 7 calculates the moving distance of an object from the peak position of the function and the table of correction data registered in a ROM 8 in advance and a result is displayed 9. The operations of a buffer amplifier 14 and the circuits 2 and 4 are controlled by the signal of a timing generator 6.

Description

Detailed Description of the Invention (Field of Industrial Application) The present invention provides a speckle length measuring needle that measures the amount of movement of an object by using a speckle pattern generated by diffused light from the surface of an object irradiated with a laser beam. It is related to.

(Prior Art) Conventionally, a method has been proposed for measuring minute deformations of objects by applying speckle patterns (Japanese Patent Publication No. 59-52
963).

In this measurement method, a laser beam is irradiated onto the surface of a deformed object to be measured, an image sensor is placed in the diffusely reflected light, and based on the image signal output from the sensor, the object is deformed before and after deformation. The cross-correlation function of the output signal is calculated, and the object movement amount is measured by making the peak position of the cross-correlation function correspond to the movement amount of the speckle pattern.

(Problem to be Solved) However, in the above measurement method, the waveform of the image signal, which fluctuates rapidly as shown in Figure 4, is directly A/D converted to calculate the cross-correlation function. The number of samples was extremely large, and the calculation process to obtain the cross-correlation function from these sampling data took time, making real-time measurement difficult.

Therefore, it has been proposed that the image signal be binarized at a predetermined threshold level and a so-called polarity correlation be taken to reduce the calculation processing time (for example, "The Needs and Seeds of Optical Measurement" published by Corona). ” p. 40).

However, when polar correlation by binarization is adopted, the correlation function has a continuous distribution as shown in FIGS.
It is obtained as a set of a plurality of discrete correlation value data according to the array pitch of

Therefore, the target object has the pixel array pitch (for example, 13 μm).
m), the cross-correlation function of the speckle pattern before and after the movement is as shown in Figures 5(a) and (b), even if the moving distance is different within the range.
As shown in , the peak positions Pp of the correlation value data are the same.

In other words, in the conventional speckle length measuring meter, the image sensor
It was not possible to obtain a resolution higher than that of , and there was a limit to measurement accuracy.

An object of the present invention is to provide a speckle length measuring meter that can obtain a measurement resolution that exceeds that of a -dimensional image sensor.

Another object of the present invention is to provide a speckle length measuring meter that can perform processing for calculating moving distance at high speed.

(Means for Solving the Problems) A speckle length measuring meter according to the present invention includes a laser generator that emits a laser beam toward the surface of a moving object to be measured, and a laser beam emitting device that emits a laser beam toward the surface of a moving object to be measured. An image sensor (13) that converts the speckle pattern appearing on the opposing observation surface into an image signal, and a second image sensor that binarizes the image signal output from the image sensor (13).
A digitization circuit (4) calculates a cross-correlation function between the speckle pattern at the reference position of the target object and the speckle pattern at the subsequent movement position based on the output data of the binarization circuit (4) using a plurality of correlation values. Data (C1, C2,...C
a correlator (5) that calculates as a discrete distribution consisting of n);
Correlation value data (C1, C2,...
Cn), and a display device (9) that displays the calculation results of the circuit.

The information processing circuit extracts, from the set of correlation value data (C1, C2,...Cn),
a first means for searching for correlation value data Cp at a peak position; and a theoretical peak when a cross-correlation function is approximated as a continuous distribution from the correlation value data Cp at the peak position and two correlation value data Cf and Ca before and after it. and a second means for determining the position and calculating an object movement amount based on the theoretical peak position, and the second means of the information processing circuit is configured to: difference (Cp-
Cf) and (Cp-Ca); and the two difference data (Cp-Cf) and (Cp-Ca).
), in which correction data corresponding to the difference between the actual peak position and the theoretical peak position of the correlation value data is stored in advance at an address defined by the difference data (Cp-Cf); (Cp - Ca); and a second calculation means that calculates the moving distance of the target object based on the actual peak position and correction data. I can do it.

(Function) As the target object moves, -dimensional image sensor (1
The image signal photoelectrically converted in step 3) changes at a period corresponding to the scanning period of the speckle pattern by the image sensor (13), and is sent to the binarization circuit (4) for each speckle pattern. Then, it is binarized.

Based on the binarized data D, the correlator (5) extracts binarized data of the speckle pattern at the reference position;
A cross-correlation function of the binarized data of the speckle pattern after the object is moved is calculated. The cross-correlation function is obtained as a discrete distribution consisting of a plurality of correlation value data (C1, C2, . . . Cn) (see FIG. 5).

The calculated correlation value data is sent to an information processing circuit, and a first means of the circuit processes the correlation value data (C1, C2,...C
Search for the actual peak position pp that takes the maximum value from n).

Here, assuming a theoretical mutual + inter-eye function when the resolution of the -dimensional image sensor is brought as close to zero as possible, the correlation value data C between the actual peak position and the positions before and after it
Since p, Cf, and Ca are considered to be plotted on the distribution of the theoretical cross-correlation function, based on these data, the peak position of the theoretical cross-correlation function (
The theoretical peak position) can be determined.

The second means of the information processing circuit determines the theoretical peak position using a predetermined method, and further calculates the object movement distance based on the theoretical peak position.

The calculated travel distance is displayed on the display (9) in real time.

Further, when the second means of the information processing circuit is composed of the first and second calculation means, memory means and readout control means,
The memory means stores the difference between correlation value data Cp and Cf (Cp-C
f) and correlation value data C, p and CaO difference (Cp-Ca)
The correction data preset in the memory means is instantaneously recalled. Then, the actual peak position is corrected using the correction data, and the theoretical peak position is determined.

(Effects of the Invention) According to the speckle length measuring meter according to the present invention, measurement resolution exceeding that of a -dimensional image sensor can be obtained, and highly accurate length measurement is possible.

Further, by using the memory in which the correction data is registered in advance, the calculation process for calculating the moving distance can be accelerated.

(Example) Hereinafter, the specific configuration of the speckle length measuring meter according to the present invention will be described along with the drawings. It should be noted that the examples are for illustrating the present invention, and should not be construed as limiting the invention described in the claims or reducing its scope.

As shown in Figure 1, the measuring head (1
) includes a semiconductor laser (11), a collimator lens (12) for shaping the laser light from the semiconductor laser (11) into a parallel laser beam (17), and an object in a sealed casing having a beam exit window (10). It is configured by arranging a one-dimensional image sensor (13) made of a COD that photoelectrically converts a speckle pattern formed by diffuse reflection on an object (18).

FIG. 2 shows an example of the configuration of a measuring circuit connected to the measuring head (1) to calculate and display the amount of movement of an object based on the image signal from the dimensional image sensor (13). ing.

As is well known, the dimensional image sensor (13) repeats scanning in the CCD array direction at regular intervals in response to the reset signal, start signal, and shift signal sent from the buffer amplifier (14). The output signal of the sensor (13) is passed through the first stage amplifier (15) to the sample and hold circuit (
2), thereby removing noise peculiar to COD.

The output signal of the sample and hold circuit (2) is sent to the binarization circuit (4) via the gain control amplifier (3), whereby the image signal S as shown in FIG. The converted data D is sent to the correlator (5), and as described later, the cross-correlation function of the speckle pattern that changes as the target object moves is repeatedly calculated, and the results of this calculation are sent to the microcomputer (7). .

The microcomputer (7) is equipped with a ROM (8) as an external storage device, which is addressed by two address data, the upper 8 bits and the lower 8 bits, and the peak position of the cross-correlation function and the ROM (8) are previously stored. The moving distance of the object is calculated from a registered table of correction data, which will be described later, and the result is digitally displayed on the display (65).

Further, the operations of the buffer amplifier (14), sample hold circuit (2), binarization circuit (4), and correlator (5) are controlled by timing signals supplied from a timing generator (6). .

The image signal obtained from the gain control circuit (3) has a period T of the start signal sent from the timing generator (6) to the -dimensional image sensor (13), as shown in FIG. occurs every time, and the period T. The generation period T of the image signal within corresponds to one scanning time of the speckle pattern by the -dimensional image sensor (13).

The image signal is sent to the binarization circuit (4) in FIG. 2, and is binarized using the average value (L in FIG. 4) of the generation period T as a threshold level. Binarized data D
is further supplied to a correlator (5).

The correlator (5) can adopt various known configurations, for example, is composed of first and second shift registers, a plurality of EX-OR circuits, an adder, etc., and has the period T.

The cross-correlation function is calculated as follows based on the data D that is input one after another.

That is, a series of data D corresponding to one speckle pattern sent at the start of length measurement is set in the first shift register, and then a series of data D corresponding to one speckle pattern sent one after another as the object moves. are sequentially set in the second shift register, and a cross-correlation function of the data set in both shift registers is calculated. The cross-correlation function is obtained as a discrete distribution consisting of a plurality of correlation value data (C1, C2, . . . Cn) (see FIG. 5). Then, when the peak value of the cross-correlation function falls below a predetermined level, the data in the second shift register at that time is transferred to the first shift register.

Thereafter, the cross-correlation function is repeatedly calculated in the same manner as described above, using the updated data in the first shift register as a reference.

The calculated correlation value data is stored in the microcomputer (7)
The microcomputer (7)
A group of correlation value data corresponding to two specification patterns (
C1, C2,...Cn), the calculation processing procedure (third
Execute (Figure).

Here, the measurement principle according to the present invention will be explained with reference to FIG.

A group of correlation value data (C1, C2,...Cn) corresponding to one speckle pattern obtained from the correlator (5)
) is a discrete value, and as shown by the solid line in Figure 6, the resolution of the one-dimensional image sensor, that is, the CCD array pitch Q (
For example, it can be drawn as a histogram with a horizontal axis pitch of 13 μm).

If we pay attention to the data Cp at the actual peak position Pp, which has the maximum value among these correlation value data, and the two data Cf and Ca before and after it, we can see that the theoretical cross-correlation when the polar correlation by binarization is not taken. It is considered that the distribution of the function changes continuously as shown by the chain line, and the three data Cf, cp, and Ca are plotted on this continuous distribution. The theoretical peak position Z of the cross-correlation function as a continuous distribution can be estimated to be at a position separated from the actual peak position X by a correction distance Y.

Therefore, the theoretical peak position Z can be determined from the actual peak position X and the corrected distance Y of the correlation function, and this theoretical peak position Z has higher accuracy than the actual peak position. Further, when the correlation data Cf and Ca before and after the peak have the same value, the actual peak position and the theoretical peak position match.

Although various prediction methods or interpolation methods can be used to obtain the corrected distance Y, here, in consideration of simplifying data processing, the three data Cf are used.

Using only Cp and Ca as basic data, the distribution of the continuous cross-correlation function is approximated by straight lines on the left and right sides of an isosceles triangle with the theoretical peak position as its apex, and the correction amount is determined by interpolation.

In this case, the ratio Y/Q between the correction distance Y and the CCD arrangement pitch Q can be determined by the following equation.

(I) When Cf<Ca, Q2X(Cp-Cf) (n) When Cf=Ca, 1=0...(2) (III) When Cf>Ca, Q2x(Cp-Ca) In Figures 5(a) to (d), there is no change in the peak position Pp, but in Figure 5(e), the peak position Pp is CCDC.
The C array pitch Q has been moved, and in this case, the ratio Y/Q is calculated using correlation value data with the peak position Pp' after the movement as the new peak position.

Furthermore, in this embodiment, in consideration of the overall efficiency of a series of arithmetic processing processes to be described later, the number of divisions M (for example, 16) when the CCD arrangement pitch is further divided and interpolated is adjusted to the ratio Y/Q. The product V c =MxY/Q - (4) is used as the interpolated value Vc in the calculation process.

Therefore, with the actual peak position pp at the start of length measurement as the origin,
Immediately after that, the actual peak position Pp when the correlation value data Cf5Ca before and after the peak becomes the same is 1, and then the actual peak position P when the correlation value data Cf and Ca become the same
If p is 2,... and the value of the actual peak position pp that occurs at the interval of the CCD array pitch is an integer value, and if the moving direction is reversed, Pp is expressed as a negative value, then the object movement at any point in time is The distance R is expressed by the following formula.

R=(PpXM+Vc)XQ/M=・(5) Next, the contents of the ROM (8) shown in FIG. 2 will be explained.

The ROM (8) stores the difference between correlation value data Cp and Cf (Cp
-Cf) is the upper 8 bits of address data AH1 correlation value data Cp and CaO difference (Cp-Ca) is the lower 8 bits of address data AL, and based on these difference data, the equations (1) to (4) are calculated. The interpolated value Vc obtained is
All possible difference data is calculated and registered in advance.

Therefore, according to the address data AH and AL, RO
The interpolated value Vc can be called instantaneously from M(8),
This reduces calculation processing time.

In addition, the ROM (8) by the microcomputer (7)
) is well known, so the explanation will be omitted.

FIG. 3 shows the processing procedure of the microcomputer. First, a group of correlation value data (C1, C2,...Cn) corresponding to one speckle pattern is taken in from the correlator, and the peak position Pp is calculated from these data. Explore. Next, the difference data (Cp-Cf) and (Cp-Ca) are calculated from the peak position and the correlation value data before and after the peak position, and these are output to the ROM (8) as address data. Then, the interpolated value Vc of the corresponding address is read from the ROM (8), and the object movement distance R is calculated from the value X of the peak position Pp and the interpolated value Vc based on the above formula (5), and the result is displayed on the display. (9) is output.

On the display (9), the moving distance R is displayed one after another in real time as the object moves.

According to the above-mentioned speckle length measuring meter, for example, when the number of interpolation divisions M is 16 and the resolution of the CCD is 13 μm, the length measurement resolution is approximately 08 μm, and extremely high accuracy can be obtained.

However, since the moving distance is calculated using a ROM table in which interpolated values are registered in advance, the calculation processing speed is increased, and even when the target object is moving at a high speed, the aforementioned high resolution can be achieved. I can do it.

The above description of the embodiments is for illustrating the present invention, and should not be construed to limit or reduce the scope of the invention described in the claims.

Further, it goes without saying that the configuration of each part of the present invention is not limited to the above-mentioned embodiments, and various modifications can be made within the technical scope of the claims.

[Brief explanation of drawings]

Fig. 1 is a perspective view showing the schematic configuration of the speckle length measuring meter, Fig. 2 is a block diagram of the measuring circuit, Fig. 3 is a flowchart showing the processing procedure of the microcomputer, Fig. 4 is a waveform diagram of the image signal, and Fig. 4 is a waveform diagram of the image signal. 5(a) to 5(e) are a series of graphs showing changes in the distribution of correlation value data, and FIG. 6 is an explanatory diagram of the measurement principle of the present invention. (1)...Measurement head (11)...Semiconductor laser (13)...-dimensional image sensor-S...image signal

Claims (1)

[Scope of Claims] [1] A laser generator that emits a laser beam toward the surface of a moving object to be measured, and a speckle pattern appearing on an observation surface facing the laser irradiation surface of the target object. An image sensor (1
3), a binarization circuit (4) that binarizes the image signal output from the image sensor (13), and a binarization circuit (4) that binarizes the image signal output from the image sensor (13); A correlator (5) that calculates a cross-correlation function between the speckle pattern and the speckle pattern at a subsequent movement position, and correlation value data (C
_1, C_2,...C_n), and a display device (9) that displays the calculation result of the circuit. (C_1, C_2,...C_n); and approximating a cross-correlation function as a continuous distribution from correlation value data Cp of the peak position and two correlation value data Cf and Ca before and after it. a second means for determining a theoretical peak position in the case where the object is moved, and calculating an object movement amount based on the theoretical peak position. [2] The second means of the information processing circuit calculates the difference (Cp-
Cf) and (Cp-Ca); and the two difference data (Cp-Cf) and (Cp-Ca).
a memory means in which correction data corresponding to the difference between the actual peak position of the correlation value data and the theoretical peak position is stored in advance at an address defined by the difference data (Cp-Cf) and ( Cp-Ca); and a second calculation means that calculates the moving distance of the target object based on the actual peak position and correction data. A speckle length measuring meter according to claim 1.