JPH0776684B2 - Optical contour measuring device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of application] The present invention uses a self-scanning one-dimensional photodiode array (hereinafter simply referred to as an image sensor) to measure in an apparatus for optically measuring the outer dimensions of a wire or the like. Regarding temperature correction of values.

[Conventional technology]

As a device for measuring the outer dimensions of a wire or the like in a non-contact manner, a device using an image sensor has been previously developed and filed in this application (Japanese Patent Application No. 58-190292).

This conventional optical contour measuring device will be described.

4 and 5 show an optical system of the optical contour measuring apparatus (1), and (2) is a light projector, which comprises a light emitting source (3), the lens (4) and a lens (5). (6) is a light receiver,
Cylindrical lens (7) and image sensor (8)
Consists of.

(9) is a wire rod which is a detected object.

Here, as the light emitting source (3), for example, a solid light emitting element such as an LED or a flash lamp is used, and the pulsed light is condensed by the rod lens (4) to form a linear light source having a uniform light intensity distribution. This linear light source is collimated by the lens (5) and enters the light receiver (6) through the wire (9). A cylindrical lens (7) in the light receiver (6) collects a component orthogonal to the collimated direction of the incident light and makes it enter the image sensor (8) to enhance the detection sensitivity.

The measurement is performed as follows.

When the light source (3) emits a pulsed light, the light with the shadow of the wire (9) enters the image sensor (8). Immediately after this, when a clock pulse is applied to the image sensor (8), as shown in FIG. 6, from each photocell in the image sensor, an output a having a shape in which short-period pulses are consecutively obtained is obtained. This output waveform a is binarized at a predetermined slice level (S, S ') of the Schmitt circuit to form a rectangular wave b. When the time points at which the falling edge and the rising edge of the rectangular wave b are generated are counted from a fixed reference time point with a predetermined clock pulse and the difference between them is calculated, the calculated value x represents the outer dimension of the wire (9).

[Problems to be solved by the invention]

The optical profile measuring device (1) is an image sensor (8).
Due to the above optical resolution, for example, when an element (cell) having a length of 7 μm per bit is used, a measurement accuracy of about 0.01 mm can be obtained in principle.

However, in actual use, an error occurs due to the temperature drift of the image sensor (8), and it is necessary to correct it by some means.

The reason why this error occurs will be described.

Looking at the output waveform a of the image sensor (8), the envelope of the portion with respect to the edge of the measured object (9) changes gently as shown in FIG. This is due to the fact that the light from the light emitting source cannot be perfectly collimated and the Fraunhofer diffraction at the edge of the measured object (9). This output waveform a is binarized depending on whether or not it exceeds a certain slice level (S, S ') of the Schmitt circuit. For example, the output waveform a of the image sensor (8) is changed by the rise of ambient temperature. When the temperature drifts in the positive voltage direction as indicated by the dotted line in FIG. 6, the time points at which the falling and rising edges of the binarized rectangular wave b change. Therefore, the length x'of the portion corresponding to the shadow of the object to be measured counted by the clock pulse is smaller than the length x measured at the reference temperature, and a measurement error occurs.

Conventionally, as a method of correcting this, the magnitude of the voltage of the output waveform a of the image sensor is monitored, and the slide level is corrected according to the fluctuation, or the slide level is corrected according to the measured temperature. There was a method to install a temperature sensor separately.

However, these conventional correction methods have the following problems.

In the floating slice method, the peak level or ground level of the output waveform a of the image sensor is detected by a sample and hold circuit, and this is divided by a predetermined voltage dividing circuit to automatically correct the slice level. However, with this method, the characteristics of the amplifier circuit, analog switch, and capacitor that make up the sample-hold circuit and voltage divider circuit change with temperature.Therefore, the design of the correction circuit must also consider these characteristic changes. Becomes difficult,
High-precision correction is impossible.

The temperature sensor can be installed by directly correcting the slice level by the output voltage of the temperature sensor, or by measuring the temperature drift characteristics in advance for each image sensor of the external shape measuring device and comparing the temperature with the correction value. There is a method of creating a digital value and storing it in the ROM of the microcomputer.

However, the former method of analog correction uses only a temperature sensor that matches the characteristics of each image sensor, so only low-accuracy correction is possible. Since a table is created for each image sensor, high-accuracy correction can be performed, but on the other hand, it requires a lot of time and labor for production, which is an inappropriate problem for mass production.

[Means for solving problems]

The present invention has been proposed in view of the above problems, and a means for solving the above problems is to irradiate an object to be measured with parallel light and project a shadow of the object onto an image sensor, and output the image sensor. Is measured by counting the falling and rising positions of a rectangular wave obtained by binarizing at a predetermined slice level of the Schmitt circuit by a predetermined clock pulse, and the difference between the measured values determines the outline of the measured object. When measuring the dimensions, in the device that corrects the measured value of the measured temperature with respect to the reference temperature, measure the light shields for the fall detection and the rise detection at the proximity positions in front of both ends of the image sensor. It is fixedly arranged so as to slightly project from the end into the measurement range, and is based on the falling and rising edges of the rectangular wave generated by the edge of each light shield. Calculate the difference between the values measured at sub-temperature and those measured at actual measurement temperature,
It is characterized in that the falling and rising positions of the rectangular wave obtained when the object to be measured is measured by these are individually corrected to perform the measurement.

[Action]

According to the above means, fluctuations due to temperature drift of the measured values of the falling and rising edges of the output waveform formed by the shadow of the edge of the detected object, the measured value of the edge of the light shield measured under the same conditions as that Since the correction value is individually corrected by the variation value from the reference temperature to the actual measurement temperature, even if the actual measurement temperature fluctuates in a large range, the measured value when the object to be detected is measured under the reference temperature can be obtained.

In particular, since the present invention corrects the measured values of the falling edge and the rising edge individually, it is sufficient that the two light shields can form the shadow of the edge, and their fixed position and the distance between them do not matter. Therefore, the two light shields can be fixedly arranged at the positions in front of both ends of the image sensor so as to slightly project from the measurement end into the measurement range.
A wide measurement range can be secured by effectively using the detection area of the image sensor, and the degree of freedom of putting in and out the object to be measured can be increased with the front space left open.

〔Example〕

 An embodiment of the present invention will be described below with reference to the drawings.

1 and 2 show an optical profile measuring device (10) of the present invention.
Shows the optical system of. The difference between this optical system and the conventional one (1) shown in FIGS. 4 and 5 is that an image sensor (8)
The light shields (11) (11) are arranged in front of, so as to sandwich the measurement range from both sides. Since the other components are the same, the same reference numerals are given and the description is omitted.

The light shields (11) and (11) may be such that their edges cause the image sensor (8) to project a light shadow similar to that of the edge of the object to be detected (9). Since these two light shields (11) and (11) may form the shadow of the edge, as shown in FIGS. 1 and 2, the measurement ends are measured at the close positions in front of both ends of the image sensor. It is fixedly arranged so as to slightly project within the range.

In the above optical system, when the light emitting source (3) emits a pulsed light, the image sensor (8) receives the parallel light shadowed by the object (9) and the light shields (11, 11). When this clock pulse is directly applied, an output A having a waveform as shown in FIG. 3 is obtained from the image sensor (8).

Here, the present invention corrects the temperature drift of the image sensor (8) by the method described below.

The waveform indicated by the solid line in FIG. 3 is obtained at the reference temperature (for example, 20 ° C.), and the waveform indicated by the dotted line is obtained at the actually measured temperature higher than that. Both the peak level and the ground level of the waveform indicated by the dotted line are higher than the waveform indicated by the solid line due to the temperature drift of the image sensor (8).

Therefore, in the rectangular wave B obtained by binarizing the output waveform at the time of actual measurement shown by the dotted line at the predetermined slide level of the Schmitt circuit, the length x ′ of the portion corresponding to the shadow of the detected object (9) is It is smaller than the length x at the reference temperature.

Thus, in FIG. 3, the magnitudes of deviations d and u due to temperature drift of the measured values of the falling and rising of the output waveform A generated by the edges of the light shields (11) and (11).
Then, the magnitudes of deviations d ′ and u ′ of the measured values of the falling and rising of the output waveform generated by the edge of the detected object (9) due to temperature drift are equal to each other.

The reason for this is that the edge shapes and installation positions of the light shields (11) and (11) are adjusted as described above, so that the shapes of the rising and falling envelopes of the output waveform A of the image sensor are The shield (11) (11) and the object to be measured (9) are the same, and the Schmitt circuit provided with hysteresis characteristics to prevent ringing has a slice level S for rising detection, This is because the fall detection slice level S ′ and the slice level S ′ are constant even if the temperature changes.

Therefore, in the device of the present invention, the values m and n measured at the reference temperature (for example, 20 ° C.) at the falling and rising of the waveform wave B generated by the edges of the light shields (11) (11) that are fixedly arranged are The measured values m ′ and n ′ of the falling and rising edges of the rectangular wave generated by the edges of the light shields (11) and (11), which are stored in advance and are actually measured, are stored in the memory. From the difference between the values m'and n ', the deviation magnitude d = m-m' of the fall measurement value and the deviation magnitude u = n-n 'of the lowland on the rising side are obtained as correction values, respectively. Next, these correction values d and u are respectively added to the falling measurement value o and the rising measurement value p of the rectangular wave generated by the edge of the measured object (9).

The measured value (O = o + d) corrected in this way (P = p +
If the difference | O-P | of u) is calculated, the calculated value | O-P | becomes the same as the value x obtained by measuring the outer dimension of the detected object (9) at the reference temperature.

〔The invention's effect〕

The present invention is based on the idea that the measurement error caused by the temperature drift of the image sensor is individually corrected for the rising and falling detection positions, and the two light shields are provided in the vicinity of the front ends of the image sensor to widen the detection area. It is fixedly placed while keeping it. By disposing this light shield, it is possible to effectively use the detection area of the image sensor and ensure the open state of the measurement space. This invention does not add an electronic component that causes a new temperature characteristic variation for correction,
Since the temperature is corrected according to the characteristics of the image sensor itself used, the temperature can be corrected with high accuracy without increasing the manufacturing cost. Incidentally, in the measurement data measured by the device of the present invention, the error is within ± 1 bit in the range of −10 ° C. to 50 ° C., and the measured value is ± 0.01 mm.
The following measurement accuracy was obtained. The error is ± 7 bits in the range of -10 ℃ to 50 ℃ when the temperature is not corrected at all, and the error in the floating slice method or the analog correction method using the temperature sensor is 0 to 40 ℃. It can be seen that the correction is performed with extremely high accuracy, compared with the range of ± 2 bits.

[Brief description of drawings]

1 to 3 show an embodiment of the present invention. FIGS. 1 and 2 are front and plan views of an optical system, and FIG. 3 is an output waveform A of an image sensor and binarization thereof. It is a correspondence diagram of the waveform wave B obtained by. 4 to 6 show a conventional example, FIGS. 4 and 5 are front and plan views of an optical system,
FIG. 6 is an inter-correspondence diagram of the output waveform a of the image sensor and the rectangular wave b obtained by binarizing the output waveform a. (2) …… Emitter, (6) …… Receiver, (8) …… Image sensor, (9) …… Detected object, (10) …… Optical external measuring device, (11) …… Light shield Body, (A) ... output waveform of image sensor, (B) ... rectangular wave, (d) ... deviation of fall detection position, (u) ... deviation of rise detection position.

Claims (1)

[Claims] 1. A rectangular wave obtained by irradiating an object to be measured with parallel light, projecting a shadow of the object onto an image sensor, and binarizing an output of the image sensor at a predetermined slice level of a Schmitt circuit. Measurement is performed by counting the falling and rising positions by a predetermined clock pulse, and when measuring the external dimensions of the measured object by the difference in the measured values, the correction is made by the amount of change in the measured temperature value with respect to the reference temperature. In the device to perform, in the vicinity of the front of both ends of the image sensor, the light shields for fall detection and rise detection are fixedly arranged so as to slightly project from the measurement end into the measurement range. The measured values of the falling and rising edges of the rectangular wave generated by the edge at the reference temperature and those measured at the actual measurement temperature Calculate the difference between
An optical contour measuring device characterized in that the falling and rising positions of a rectangular wave obtained when measuring an object to be measured by these are individually corrected to perform the measurement.