JP3149632B2 - Cross section shape recognition device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a section for recognizing a sectional shape or a contour of an object by continuously measuring the distance to each position of the object using the principle of triangulation. The present invention relates to a shape recognition device.

[0002]

2. Description of the Related Art For example, FIG.
There is something like that shown in FIG. 8 is a block diagram showing a control unit of a conventional cross-sectional shape recognition device. As shown in FIG. 1, the cross-sectional shape recognition apparatus includes a spot light source 1 that irradiates a spot light Ls to an object to form an irradiation image on the object. The spot light source 1 is a semiconductor laser, and is connected to a semiconductor laser driving unit 2. One-dimensional C for receiving reflected light Lr from the irradiation image
It has a CD sensor 3. This one-dimensional CCD sensor 3
A lens 4 for condensing the reflected light Lr is attached to the front surface of. As shown in FIG. 3, the one-dimensional CCD sensor 3
Has a light receiving portion 3a composed of photodiodes, that is, pixels 5, aligned in a straight line, and the pixels 5 receive reflected light Lr from an irradiation image passing through the lens 4. The direction of the irradiation image is detected based on the position of the pixel 5 that has received the reflected light Lr. Each pixel 5 detects the amount of received light based on the amount of charge generated according to the amount of received light within a predetermined time. Thus, when the direction of the irradiation image with respect to the one-dimensional CCD sensor 3 is detected, the distance to the irradiation image can be calculated based on the angle calculated from this direction by using the principle of triangulation. This calculation method is described in, for example, Japanese Patent Application No. Hei.
This method is disclosed in the detailed description of the invention of Japanese Patent Application Laid-Open No. 31986 (Japanese Patent Application Laid-Open No. 6-241784).

The cross-sectional shape recognition device has a mirror 7 (reflecting means) rotatably supported and rotationally driven by a motor 6 (driving means). An encoder 8 that outputs a pulse signal indicating the amount of rotation of the motor 6 is connected to the rotation shaft of the motor. The mirror 7 irradiates the spot light Ls from the spot light source 1 toward the object to be measured, and reflects reflected light Lr from an irradiation image formed on the object to be measured (not shown) to the one-dimensional CCD sensor 3. When the mirror 7 is rotated, the object to be measured can be scanned by the spot light Ls. Therefore,
The cross-sectional shape can be recognized by continuously measuring the distance to the irradiation image while scanning the object to be measured with the spot light Ls.

[0004] However, in the above-mentioned irradiation image on the object to be measured, generally, the boundary between a portion irradiated with light and a portion not irradiated with light is not clear. As described above, when reflected light or the like is received, the light receiving range has an extension. Therefore, the amount of light received by each pixel 5 corresponds to the luminance of the irradiated image as shown in the graph of FIG. Distribution.

Therefore, conventionally, for example, the irradiation direction is determined from the position of the pixel having the highest light receiving intensity, or the following equation (1) is obtained by using the position of each pixel and the light receiving intensity at each pixel. , The value K indicating the position of the center of gravity of the light, that is, the position where the amount of irradiated light is balanced on both sides, is determined, and this position is regarded as the light receiving position to determine the irradiation direction. Note that i is a value indicating the bit position of the pixel of the one-dimensional sensor, and vi is an output value of the i-th pixel of the one-dimensional sensor.

[0006]

(Equation 1)

Here, a process for determining the position of the center of gravity of the light and determining the direction of the irradiation image will be described. When the motor 6 for driving the mirror 7 is operated, the pulse signal of the encoder 8 connected to the rotation shaft of the motor 6 is counted by the counter 9, and the irradiation direction of the spot light Ls is stored in the memory 10. When the irradiation direction of the spot light Ls is detected, the one-dimensional CCD sensor 3 receives reflected light from an irradiation image formed by irradiation with the spot light Ls, and outputs a measured value of the received light amount to the video buffer 11. I do. The measured value output to the video buffer 11 is simultaneously the A / D of each pixel 5
The data is converted into a numerical value indicating the measured value of the amount of received light by the converter 12, and is sequentially stored in the memory 10. When the digitization of the measurement data of all the pixels 5 is completed, the irradiation direction data stored in the memory 10, that is, the angle and the digitized measured value of the received light amount are read into the CPU 13, and the optical barycenter is calculated by Expression (1). Is calculated.

The distance to the irradiation image is calculated from the position of the optical center of gravity and a conversion table (not shown) for converting the position of the optical center of gravity and the distance prepared in advance. The conversion step and the conversion table are disclosed, for example, in the detailed description of the invention in Japanese Patent Application No. 5-31986. Such processing,
The cross-sectional shape is recognized by continuously performing the scanning with the spot light on the object to be measured.

[0009]

However, the pixel 5 for detecting the amount of received light based on the amount of electric charge generated in accordance with the amount of received light usually has a limited electric charge accommodating capacity. When light having a capacity higher than the capacity is generated by receiving the light, a so-called charge saturation phenomenon occurs, and charges that cannot be stored overflow. In this case, the measured value of the pixel does not indicate the actual amount of received light.

If such a defect occurs in, for example, a pixel receiving the largest amount of light or a pixel adjacent thereto,
A graph showing the distribution of the received light amount of each pixel is as shown in FIG. 5, and the position of the pixel having the largest received light amount and the measured value of the received light amount at that pixel cannot be detected. Further, from such a graph, it is not possible to calculate an accurate position of the optical barycenter. Such a problem can be avoided by reducing the amount of generated charge per unit received light amount. However, if the amount of generated charge per unit received light amount is too small, the measurement accuracy is reduced. Therefore, a method of determining the amount of generated electric charge per unit amount of received light for each measurement is also conceivable. However, such a problem cannot always be prevented.

The inventor of the present invention focuses on the fact that the shape of the graph of the measured value of the amount of received light measured in successive distance measurement operations, that is, the tendency of increase and decrease, is extremely similar. Reached. The cross-sectional shape recognition device according to the present invention, when the output value of the light reception amount measurement of some pixels is not a value indicating the actual light reception amount, the measurement value of some pixels that do not indicate the actual light reception amount Is corrected to a value corresponding to the actual measured value using the measured value of the received light amount in the past, thereby solving the above problem.

[0012]

According to the present invention, there is provided a cross-sectional shape recognition apparatus for irradiating a spot light on an object to form an irradiation image on the object. A light receiving unit comprising a plurality of light receiving members for measuring the amount of light received from the irradiated image, and a reflecting unit for guiding the spot light to the object to be measured and guiding the light from the irradiated image to the light receiving unit, A driving unit for rotating the reflection unit to scan the spot light, and calculating a center of gravity of the irradiation image based on a measured value of the amount of light received by the plurality of light receiving members, and calculating a distance to the center of gravity. A cross-sectional shape recognition device having a calculating means, a storage means for storing a measured value of a past light reception amount already measured by the cross-sectional shape recognition device, and a part of the plurality of light receiving members being measurable. The light outside the range measurable by receiving
The limit value of the appropriate range is output as a measured value and the actual
If not shown, a value indicating the actual amount of received light is calculated based on the past measured values stored by the storage means.
And a correction means for performing correction .

[0013]

When the light receiving member receives light exceeding the measurable range, the light receiving member outputs a measured value that does not indicate the actual amount of received light. In such a case, the measured value that does not indicate the actual received light amount is corrected to a value corresponding to the actual received light amount using the past measured value.

[0014]

Next, an embodiment of the present invention will be described in detail with reference to the drawings. The members already described are given the same reference numerals. FIG. 1 is a conceptual diagram showing a structure of a sensor unit 20 of a cross-sectional shape recognition apparatus of the present embodiment, that is, a portion including a spot light source 1, a one-dimensional CCD sensor 3, and a mirror 8. The structure of this part is the same as that of the conventional device. FIG. 2 is a block diagram showing a control unit for calculating the distance from the various measurement values of the one-dimensional CCD sensor 3 to the position to be measured, and a part of the sensor unit 20.

As shown in FIG. 1, a spot light source 1 for irradiating a spot light is provided on a sensor section 20 of the sectional shape recognition apparatus.
Is attached. The irradiation port of the spot light source 1 can be freely changed in diameter, and is appropriately selected depending on the surface condition of the object to be measured, measurement accuracy, and the like. A mirror 7 that reflects the spot light Ls from the spot light source 1 toward the object to be measured is rotatably attached to the sensor unit 20, and is driven by a motor 6. Therefore, by rotating the mirror 7 by the motor 6, the object to be measured can be scanned with the spot light Ls. Note that distance measurement is performed every few milliseconds when an object having a relatively smooth surface with a small surface step is used as the object to be measured. An encoder 8 for generating a pulse signal indicating the rotational position of the motor 6 is connected to a drive shaft of the motor 6. Therefore, encoder 8
By detecting the position of the motor 6 by counting the rotation drive signal from the counter 9 by the counter 9, the position of the mirror 7 can be detected and the irradiation direction of the spot light Ls can be detected. Note that the irradiation direction of the detected spot light Ls is stored in the memory 10.

The sensor section 20 further includes a one-dimensional CCD for receiving light from an irradiation image of the spot light Ls on the object to be measured.
The sensor 3 is attached. As shown in the figure, a lens 4 for passing reflected light Lr from a measurement target such as an irradiation image of a spot light Ls is attached to the front surface of the one-dimensional CCD sensor 3, and the reflected light Lr passing through the lens 4 is attached. Is provided. As shown in FIG. 3, the light receiving section 3a is composed of a number of pixels 5 arranged in a straight line, and reflected light Lr from an irradiation image to be measured.
The irradiation direction of the received light is detected based on the position of the pixel 5 that has received the light. In addition, in consideration of the shift of the light receiving position in the width direction, a wide pixel may be used, or the number of pixels may be two or more. The measurement data of the amount of received light measured by each pixel 5 of the one-dimensional CCD sensor 3 is output as image data to the video buffer 11 shown in FIG. The measurement data output to the video buffer 11 is simultaneously converted into a numerical value by the A / D converter 12 for each pixel 5,
It is sequentially stored in the memory 10.

As shown in FIG. 2, the memory 10 is connected to a CPU 13, a ROM 22, a RAM 23 via a bus 21.
And the register 24, so that the digitized measured values stored in the memory 10 can be freely transferred. That is, the measured value stored in the memory 10 can be read into the CPU 13, the numerical value calculated in the middle of the calculation can be written into the register 24, and the numerical value in the register 24 can be read. CPU
Numeral 13 calculates the position of the optical center of gravity from the measured value of the amount of received light. In the present embodiment, it also serves as a correction means for performing a comparison step and a correction step described later.

The bus 21 has a past video buffer memory 31 for storing measurement data measured in the past by the one-dimensional CCD sensor 3, and a numerical value such as the position of the optical center of gravity calculated from the past measurement values. Is stored in the memory 32 for storing past numerical values. Therefore, a measured value that does not indicate the actual amount of received light can be corrected to a value corresponding to the actual amount of received light using the past measured values and numerical values indicating the position of the optical center of gravity.

The operation flow when measuring the cross-sectional shape of the object to be measured by such a cross-sectional shape recognition device will be described.
First, a spot light L is directed from a spot light source 1 to an object to be measured.
s, and an irradiation image of the spot light Ls is formed on the object to be measured. Then, the reflected light Lr from the irradiation image is converted into a one-dimensional CC.
The light is received by the D sensor 3. As shown in FIG.
The D sensor 3 has a number of pixels 5, and measurement data of the amount of light received at each pixel 5 is stored in the video buffer 11. A set of measurement values in one measurement, that is, a measurement data group is simultaneously digitized for each pixel 5 by an A / D converter, and is stored in the memory 10 as one measurement value group.
Is stored.

Next, it is determined whether or not any of the measured values of the measured value group needs to be corrected. This determination is made based on whether or not a measured value indicating the maximum value of the measurable range exists as the measured value of the pixel 5. This is because when a pixel whose measured value is the maximum value of the measurable range actually receives more light, the measured value needs to be corrected.

At this time, if there is no value indicating the maximum value of the measurable range and the distribution of the actual amount of received light of each pixel 5 is measured over the entirety as shown in FIG. There is no need to correct the value. Therefore, the position of each pixel 5 and the measured value of the amount of received light at that position are substituted into the above equation (1) to calculate the position of the optical center of gravity, and the irradiation image determined based on this is calculated. And the principle of triangulation, the distance to the irradiation image is calculated. On the other hand, when there is a value indicating the maximum value of the measurable range, that is, as shown in FIG.
If the position where the amount of received light is maximum cannot be read from the graph, it is necessary to correct a measured value that does not indicate the amount of received light to a value corresponding to the amount of received light.

The steps are as follows. In the measured value group, a measured value indicating the maximum value of the measurable range of the pixel 5 exists, and when it is determined that the correction is necessary, the measured value of the measured value group and the corrected value are unnecessary. The measured values of the past measured value group are compared with the measured values. As the past measurement value group used for comparison, a measurement value group in the latest measurement, that is, a measurement performed several milliseconds ago is usually used. In this comparison, the graph of the measured value group being processed and the graph of the past measured value group are superimposed at a position where the values on the vertical axis, that is, the measured values of the received light amount increase and decrease coincide with each other. The purpose is.

The comparison process for performing the comparison will be specifically described with reference to a comparison routine shown in FIG. 7. First, a graph G1 of a measured value group during processing and a graph G2 of a past measured value group are shown in FIG. As shown in (1), drawing is performed on the same sheet (S1).
At this time, the graphs G1 and G2 are relatively parallelly moved in the Y-axis direction so that the measured light-receiving levels of the pixels 5 that do not receive light coincide with each other.

Next, regarding the graph of the measured value during processing,
The sum Sf of the amount of light received from a given pixel (In ₁₎ to In ₂ -th pixels with calculated by Equation (2), the received light amount of the pixel from the ₃ th In to a predetermined position of the pixel (In ₄₎ Is obtained by equation (3) (S2). Similarly,
For the graph of the past measurement value, the specified pixel (In ₁ )
From the received light amount of the sum S f'up In ₂ -th pixels with calculated by Equation (4), the sum of the received light amount of the pixel from the ₃ th In to a predetermined position of the pixel (In ₄₎ S b' Is obtained by equation (5) (S3).

[0025]

(Equation 2)

When obtaining each sum, the measured value level previously matched is set as a reference, that is, 0. Also In ₁
Figures -In ₄ is a numerical value indicating the pixel position using the convenience of explanation in the comparison step, it does not represent the actual pixel positions at the time of measurement. The numerical value calculated in this way is substituted into the equation (6) to obtain an evaluation function H (S4). The comparison operation including the series of steps (S1 to S4) is performed each time the graphs are relatively translated in the X direction one pixel at a time, and a superimposed position of the graph with the smallest evaluation function is obtained. Note that the steps described here are steps for easily understanding the principle, and it goes without saying that other methods based on the same principle may be used. Further, the evaluation method for determining the optimum superposition position is not limited to the method based on this principle. For example, the measurement value of each graph having the same X-axis value, that is, in this embodiment, the pixel value The measured value V ′ during the processing at the position In is replaced with the past measured value V ″.
The work of dividing by the pixels from In ₁ to In ₂ and the pixels from In ₃ to In _{4 is} performed, and the position where the standard deviation of the obtained numerical value group is the smallest is set as the most suitable overlapping position, and so on. In addition, various methods can be used.

When the optimum superposition position is determined in the comparison step, a measured value for which the actual received light amount is not indicated is calculated based on the ratio at the optimum superposition position and the past measured value.
Correction is made to a value corresponding to the actual amount of received light (correction step). That is, v ′ on the two-dot chain line is obtained by multiplying v ″ which is a past measurement value by H.

When the measured value is corrected in this manner, the amount of light received as indicated by the two-dot chain line in FIG. 6 can be used when determining the position of the optical center of gravity. If the amount of light received in this portion is used, a more accurate optical barycenter can be obtained. That is, for example, when the region P indicated by the two-dot chain line is not included, the measured value of the amount of received light in the pixels from In ₂ to In ₃ is constant. However, in practice, as shown by a two-dot chain line, the change in the amount of received light in that section may be very large. In such a case, when obtaining the position of the optical center of gravity, the amount of light received in the region P indicated by the two-dot chain line is most important. In addition, when a measurement error M as shown by a broken line in FIG. 5 occurs, the influence of the error M on the calculation of the position of the optical center of gravity includes an area P indicated by a two-dot chain line. It depends greatly on whether or not.

When the correction of the measured value during the processing is completed in this way, the pixel position and the measured value of the amount of received light at the pixel are substituted into the equation (1), as in the case where the correction is not required. Then, the position of the optical center of gravity is calculated. Then, the distance to the irradiation image is calculated based on the direction of the irradiation image determined based on this and the principle of triangulation.

When the position of the optical center of gravity is calculated, the distance to the irradiation image is calculated from the position of the optical center of gravity and a conversion table of the optical center of gravity position and the distance prepared in advance. Such a series of operations is continuously performed while scanning the object to be measured with the spot light, and the cross-sectional shape is recognized.

As described above, according to the present invention, when some of the light receiving members receive light exceeding the measurable range and a value that does not indicate the actual received light amount is output as a measured value,
The actual measured value is calculated and supplemented based on past measured values.
Since positive, it is possible to measure the position of the optical center of gravity more accurately. Therefore, the distance between the measured positions can be measured more accurately.

[Brief description of the drawings]

FIG. 1 is a schematic plan view showing a cross-sectional shape recognition device.

FIG. 2 is a block diagram illustrating a control unit of the cross-sectional shape recognition device according to the embodiment.

FIG. 3 is a plan view showing a light receiving section of the one-dimensional CCD sensor.

FIG. 4 is a diagram showing a graph drawn from measured values of the amount of received light.

FIG. 5 is another diagram showing a graph drawn from measured values of the amount of received light.

FIG. 6 is a diagram showing a state where two graphs are drawn in an overlapping manner.

FIG. 7 is a flowchart illustrating a comparison step.

FIG. 8 is a block diagram showing a control unit of a conventional cross-sectional shape recognition device.

[Explanation of symbols]

1: spot light source, 3: one-dimensional CC
D sensor, 8 ... Mirror, 10 ...
Memory, 13 ... CPU, 31 ...
Past video buffer memory, 32... Past numeric data memory.

Claims (1)

(57) [Claims] 1. A spot light source for irradiating a spot light on an object to form an irradiation image on the object, and a light receiving means comprising a plurality of light receiving members for measuring an amount of light received from the irradiation image. Reflecting means for guiding the spot light to the object to be measured and guiding light from the irradiation image to the light receiving means,
A driving unit for rotating the reflection unit to scan the spot light, and calculating a center of gravity of the irradiation image based on a measured value of the amount of light received by the plurality of light receiving members, and calculating a distance to the center of gravity. A cross-sectional shape recognizing device having a calculating means, a storage means for storing a measured value of a past light reception amount already measured by the cross-sectional shape recognizing device, and a part of the plurality of light receiving members having a measurable range. The limit value of the range that can be measured by receiving light exceeding
Output and does not indicate the actual received light amount,
A correction means for calculating and correcting a value indicating the actual amount of received light based on the past measured values stored in the storage means .