JPS60135704A - Pattern detector - Google Patents

Abstract

PURPOSE:To detect exactly the shape of a pattern to be detected at a high speed withoug being affected by the specular reflected light from the pattern surface by synthesizing the optical images of the pattern illuminated from two directions. CONSTITUTION:A sheet 19 is alternately illuminated from two differente directions by two sets of illuminating systems consisting of halogen lamps 21a, 21b, filters 22a, 22b, condenser lenses 23a, 23b and shutters 24a, 24b. The image of the pattern of the sheet 19 is picked up by a TV camera 26 and is converted to an image signal. The image signal from the camera 26 is dinary coded by shading correcting circuits 27a, 27b and binary coding circuits 28a, 28b and thereafter the signals are synthesized by a memory ciruit 29 and an AND circuit 31.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention relates to a pattern detection device, and is particularly suitable for detecting a pattern formed by an aggregation of fine particles such as metal particles having one surface that specularly reflects light. The present invention relates to a suitable pattern detection device.

[Background of the invention]

As conventional pattern detection devices, there are known devices that use bright field (epi-illumination) illumination, devices that use dark field (oblique) illumination, devices that use transmitted illumination, and the like.

As shown in FIG. 1, a pattern detection device using bright field illumination bends the original optical path emitted from an illumination system 1 using a half mirror 2, and irradiates the pattern 3 from almost perpendicularly above. The detection system 4 similarly detects the shape of the pattern 3 from vertically above. By adopting such a configuration, it is possible to substantially match the illumination direction and the detection direction, and a bright optical image of pattern 3 can be obtained. Generally, the signal-to-noise ratio (S/N ratio) of a photoelectric conversion element such as a two-dimensional image sensor or linear sensor used in detection system 1 is the same as the amount of incident light and the detection of one screen (one line in the case of a linear sensor). It is proportional to the product of the time required. Therefore, according to a pattern detection apparatus using bright field illumination, the amount of light incident on the detection system l becomes large, so that a pattern image signal with a high S/N ratio can be obtained.

However, when a pattern detection device using bright field illumination described above detects a pattern formed by a collection of fine particles such as metal fine particles having a specular reflection surface of light, the following drawbacks occur. In other words, as shown in FIG. 2, if the pattern 3 is made of a material that is detected darker than the background 5, a part of the pattern 3 shines and becomes a bright spot even though the pattern 3 is normal. 6 may be detected. Furthermore, as shown in FIG. 3, if the pattern 3 is made of a material that is detected brighter than the background 5, a dark spot 7 may appear in a part of the pattern 3 even though the pattern 3 is normal. may occur.

The cause of the bright spots 6 and dark spots 7 described above will be explained using a pattern formed by an aggregation of fine metal particles as an example. That is, FIG. 4(a) is a cross-sectional view of the pattern 3 formed by a collection of metal fine particles 8, and FIG. 4(b) is a partially enlarged view thereof. As shown, the microparticles 8 generally have several cleavage planes 9. Therefore, Fig. 4(C)
As shown in the figure, since the illumination light 10 is regularly reflected by the cleavage plane 9 and the normal direction of the cleavage plane 9 is completely random, the reflected light also travels in completely random directions. Therefore, as shown in FIG. 2, when the pattern 3 is detected darker than the background 5, the reflected light enters the entrance pupil of the imaging lens of the detection system 4, and a bright spot appears on the pattern 3. 6 is detected. or,
As shown in FIG. 3, when the pattern 3 is detected brighter than the background 5, the point where reflected light does not enter the entrance pupil of the imaging lens of the detection system 4 is detected as a dark spot 7 on the pattern 3. be done. Therefore, these bright spots 6. The dark spots 7 form a type of square noise (hereinafter referred to as bright spot noise/dark spot noise) on the detected image, causing a decrease in the reliability of pattern detection. - As shown in FIG. 5, the pattern detection device using dark field illumination illuminates the pattern 3 from diagonally above using a parabolic concave mirror 11, etc., and the detection system 4 illuminates the pattern 3 from almost perpendicularly above the pattern 3.
Detection is performed using □. As shown in FIG. 6, when there is a step 12 between the pattern 3 and the background 5, the pattern detection device using this dark field illumination can detect only the step 12 particularly brightly, and the outline of the pattern 3. It is suitable for detecting surface irregularities. However, similar to the pattern detection device using bright field illumination, the putter 73
If it is made of a material having a specular reflection surface, such as fine metal particles, bright spot noise or dark spot noise occurs, which has the disadvantage of reducing the reliability of pattern detection.

A pattern detection device that uses transmitted illumination is known as a device that eliminates the drawbacks of the pattern detection device that uses bright-field illumination and the pattern detection device that uses dark-field illumination described above. A pattern detection device that uses transmitted illumination is
As shown in FIG. 7, an illumination system 1 illuminates the object on which a pattern 3 is formed from the back side, and a detection system 4 detects a silhouette image of the pattern 3. This pattern detection device can correctly detect the pattern shape without generating dark spot noise or bright spot noise even when the pattern 3 is formed of a material having a specular reflection surface made of metal fine particles. However, when the pattern 3 is formed on both the front and back surfaces of the object, it is difficult to distinguish between the pattern 3 on the back and the front, and the pattern shape cannot be detected correctly. In addition, if the background 5 of the pattern 3 is made of a material with low light transmittance, □ it cannot be detected on a dark silhouette image, a detection image with a high S/N ratio cannot be obtained, and the reliability of pattern detection is reduced. The disadvantage is that it decreases.

[Purpose of the invention]

The present invention was developed in view of the above-mentioned drawbacks of the prior art, and it is possible to change the shape of a pattern formed by an aggregation of fine particles such as metal fine particles having a specular reflection surface of light, without being influenced by the specular reflection light from the pattern surface. It is an object of the present invention to provide a pattern detection device that can perform accurate and high-speed detection without any interference.

[Summary of the invention]

The pattern detection device of the present invention includes an illumination means for illuminating a detected pattern on an object from two different directions, and independently detects optical images of the detected pattern illuminated from two directions, and generates two systems of image signals. The present invention is characterized in that it is comprised of an imaging means for converting the image signals into the image data, and a synthesizing means for synthesizing corresponding portions of the detected pattern of the two systems of image signals output from the imaging means.

Next, Fig. 8 (IL) l (b) # (e),
The principle of the present invention will be explained using (d). The illumination means illuminates the pattern 3 shown in FIG. 8(a) from two different directions. At this time, the reflected light at each point of the pattern 3 is detected only from a certain direction determined by the direction of the reflecting surface and the direction of illumination. Therefore, Fig. 8(b) +
As shown in (c), when the pattern 3 is illuminated from two different directions and its optical image is detected by the imaging means, the bright spot noise 6 will not appear at the same position on the detected image. Furthermore, if the numerical aperture (N, A,) of the detection optical system is set large, dark spot noise will not appear at the same position on the detection image in most cases due to illumination from two different directions. Therefore, by combining the two systems of image signals of pattern 3 illuminated from two different directions using the combining means, it becomes possible to detect a pattern free of bright spot noise and dark spot noise.

[Embodiments of the invention]

The present invention will be explained in more detail below with reference to embodiments shown in the accompanying drawings.

FIG. 9 is a diagram showing a first embodiment of the present invention.

This first embodiment is a pattern detection device particularly suitable for inspecting patterns on printed circuit boards, such as green sheet patterns. In this example, pattern 3 reflects light diffusely and
A sheet formed of fine particles of a metal (for example, tungsten or molybdenum) that is black or whose color is detected darker than the sheet 19 on a white sheet (for example, a green sheet mainly composed of alumina) 19 to be diffused; This first
This embodiment detects the shape of pattern 3 without being affected by the bright spot noise described above.

As shown in FIG. 9, the sheet 19 includes halogen lamps 21 m and 21 b connected to a light source power source, filters 22 m and 22 b, and condenser lenses 23 a and 23 b.
The light is irradiated from two different directions by two sets of illumination systems consisting of a shutter 24m and a shutter 24b. The light source is not limited to the halogen lamps 21 a and 21 b, but a normal tungsten lamp, mercury lamp, or the like may be used. The filters 22a and 22b are used for the purpose of cutting wavelength components that adversely affect the materials of the sheet 19 and the pattern 3, and wavelength components that adversely affect the aberrations and resolution of the detection optical system, which will be described later. This filter 22a.

Although it is not necessary to install the filters 22b, it is preferable to use infrared absorption filters as the filters 22m and 22b in order to avoid thermal problems caused by infrared rays and deterioration of resolution. Furthermore, a monochromatic light filter may be used to avoid a decrease in resolution due to chromatic aberration of the detection optical system. In any case, the filter 22a.

The selection of 22b is preferably determined experimentally depending on conditions such as required resolution. The shutters 24a and 24b are used to electromagnetically move a light-shielding plate to block illumination light, and are used to switch illumination from two directions in one direction.

Of course, the shutter 24a.

An electronic shutter using a liquid crystal may be used as the shutter 24b, or the shutters 24a and 24b may be omitted, and a switch may be interposed between the light source power supply silver and the halogen lamps 21 & 21b to turn on/off the halogen lamps 211L+21b.
It doesn't matter if you turn it off and switch it.

The pattern 3 illuminated from one direction by switching the shutters 24m and 24b is imaged by an imaging device consisting of an imaging lens 5 and a TV camera 26 located above the sheet 19, and converted into an image signal Ki.

Here, the imaging magnification of the lens 6 is determined by the ratio between the required microscopic resolution and the size of one pixel in TV camera %. The image signal output from the TV right camera 6 is input to the shading correction circuit υm, 27b, and sensitivity unevenness and illumination unevenness of the TV right camera 6 are removed. The shading correction circuits 27a and 27b are disclosed, for example, in "Television Signal Shading Corrector for Input to Digital Image Processing Device", Journal of the Society of Television Engineers, Vol. A, No. 5, pages 413-417 (1981). In addition to the method described above, any known method may be used.

Note that the shading correction circuits 27a and 27b may be omitted if the sensitivity unevenness and illuminance unevenness of the TV left camera described above are small enough to be ignored.

The image signals after shading output from the shading correction circuits 27m and 27b are sent to the binarization circuits 13m and 1.
3b and is binarized. Binarization circuit 13a, 1
For 3b, a method of binarizing using a fixed threshold value may be used, or a method of binarizing using a floating threshold value may be used.

As shown in FIG. 9, the shading correction circuit υ&, 27
Two systems of binarization circuits 28a and 28b are provided. Then, the shutter 24a opens and the shutter 24a opens.
If b is closed, the shading correction circuit 27
m and the binarization circuit 28a are used, and conversely, the shutter 24
When the shutter b is open and the shutter 24& is closed, the shading correction circuit 27b and the binarization circuit 28b are used. In other words, a dedicated shading correction circuit 2 is required for each case of illumination using the halogen lamp 21a and when illuminating using the halogen lamp 21b.
7a, 27b and binarization circuits 28a, 28b are used. 2
One screen worth of image signals binarized by the digitization (b) path 28a is stored in the memory 29. Then, the binarization circuit 28b
The image signal for one screen that has been binarized is read out from the memory 29 at the timing when it is output to the AND circuit 31, and both image signals are output to the AND circuit 31. The synchronization control circuit includes a shading correction circuit 27*, which is provided with two systems in accordance with the control of the shutters 24&, 24b.
27b-binarization circuit 28m, 28b, which system is used, and timing control of reading out the image signal from the memory 4 and outputting the image signal from the binarization circuit 28b. The two systems of image signals input to the AND circuit 31 are combined by the AND circuit 31.

Through the above operations, the detected image of pattern 3 shown in FIG.
Even if the beZ value conversion circuits 28m and 28b detect the bright spot 6 as shown in FIGS. 8(b) and 8(C), the synthesized detected image (AND circuit 31 output) becomes a clear image without bright spots 6.

Regarding the formation of the above composite image, FIG. 10(a),
(b) This will be explained in more detail using 1(e). FIG. 10 (IL) shows the pattern 3 (line A shown in FIG. 8 (b)
It is a time chart showing the processing state after the optical image of A' is detected as an image signal by the TV left camera. As shown, bright spot noise 61 is present in the image signal output from the TV left camera. □Also, the image signal output from the TV camera% has uneven illumination and uneven sensitivity of the TV left camera, and as shown in the figure, the shading circuit n removes this uneven lighting and uneven sensitivity. .

Then, the output of the shading correction circuit 27& is binarized by the binarization circuit 28& as shown in the figure, and read into the memory 4.

FIG. 10(b) shows that the optical image of the line BB' shown in FIG. 8(c) of the pattern 3 formed by irradiation with the halogen lamp 21b is detected as an image signal by the T'V camera 26, and then This is a time chart showing the processing status of. As shown in the figure, bright spot noise 6b is present in the image signal output from the TV left camera. This image signal is
As in the case of FIG. 10 (&), the shading correction circuit 27b removes uneven illuminance and sensitivity, and the binarization circuit 28b converts the signal into a binary signal.

When the binarized image signal is output from the binarization circuit 28b, the previously read binarized image signal is read out from the memory 4 in synchronization with this, and both image signals are The signals are synthesized by an AND circuit 31, resulting in a waveform shown in FIG. 10(e). As a result, the bright spot noises 6a and 6b are eliminated as shown in the figure.

The first embodiment described above is based on the assumption that the bright part of the screen becomes logic H in the circuit from the TV right camera 6 to the AND circuit 31, but conversely, the dark part of the screen becomes logic H. If it is assumed that HK, a NOR circuit may be used in place of the AND circuit 31.

Further, the first embodiment described above is applied to the case where the dark pattern 3 is formed on the bright background 5, but conversely, when the bright pattern 3 is formed on the dark background 5. This can be applied by replacing the AND circuit 31 in FIG. 9 with an OR circuit.

FIG. 12 is a diagram showing a second embodiment of the present invention, and FIG.
The same parts as in the first embodiment shown in the figures are given the same reference numerals, and the explanation thereof will be omitted. The differences from the first embodiment shown in FIG. 9 are as follows.

That is, the outputs of the shading correction circuits 27'' and 27b are converted into digital signals by analog/digital converters (not shown).
If a and Z7b are of the tintal type, an analog/digital converter is not required). The image signal from the shading correction circuit 27 & converted into a digital signal is
written in the memory 29 and the shading correction circuit 27
It is read out in synchronization with the output timing of b. and,
Both image signals are input to the minimum value circuit 32. The minimum value circuit 32 compares the magnitude relationship between the two input series of image signals and outputs the smaller image signal. Minimum value circuit 32
The output is binarized by binarization circuit 4 and output.

FIG. 12(a) shows the operation of this second embodiment.

(b) I ((Explain in more detail using ri. 12th
Figure (IL) shows line A shown in Figure 8 (a) of pattern 3 formed by irradiation with a halogen lamp 21 m.
It is a time chart showing the processing after detecting the optical image of A' as an image signal using a TV camera. As shown in the figure, bright spot noise 6m is present in the image signal that is the output of the TV camera. This image signal is input to the shading correction circuit 27m, and sensitivity unevenness and illuminance unevenness of the TV camera section are removed. The output of the shading correction circuit 27a is converted into a digital signal by an analog/digital converter (not shown) and then read into the memory 29.

FIG. 12(b) shows that the optical image of the line B B′ shown in FIG. 8(c) of the pattern 3 formed by irradiation with the halogen lamp 21b is detected as an image signal by the TV right camera 6, and the subsequent Time Qi showing processing? −).

Irradiation switching between the halogen lamps 21a and 21b is performed using shutters 24m and 24b. As shown in FIG. 12(a), bright spot noise 6b is present in the image signal output from the TV camera. This image signal is shown in Fig. 12(a).
Similarly to the above case, the shading correction circuit 27b removes uneven illuminance and uneven sensitivity, converts the signal into a digital signal, and outputs the signal. In synchronization with this, the image signal already read into the memory 4 is read out, and both image signals are input to the minimum value circuit 32. The minimum value circuit 32 is
As shown in FIG. 12(c), the two image signals are compared and the smaller image signal is output. Therefore, as shown in FIG. 12(c), an image signal from which the bright spot noises 6m and 6b have been removed is output. Although the image signal output from the minimum value circuit 32 is depicted as being an analog signal in FIG. 12(c), it is actually a digital signal.

Of course, a tintal/analog converter may be used to output the signal as an analog signal. The image signal outputted from the minimum value circuit 32 is inputted to the binarization circuit 4, and as shown in FIG. 12(d)
It is binarized as shown in . As is clear from the above description, according to this second embodiment, it is possible to eliminate the bright spot noises 6a and 6b with a detection system having a relatively small configuration. Furthermore, if the output of the minimum value circuit 32° in FIG. 11 is used as is, it is possible to obtain not only a 21-section image but also a multi-1u1 grayscale image.

In addition, regarding the above-mentioned second embodiment, the first actual observation 1f
Similar to p, various modifications are possible. For example, when the pattern 3 of bright gold particles is formed on the dark sheet 19, the minimum value circuit 32 may be changed to the maximum value circuit.

FIG. 13 is a diagram showing one embodiment of Brown 3 of the present invention.
The same parts as those in the first embodiment shown in FIG. 9 are given the same reference numerals, and the explanation thereof will be omitted. Like the first and second embodiments, this third embodiment is also a pattern detection device suitable for a case where a pattern 3 is formed of black metal particles on a white sheet 19 such as a green sheet pattern. be. 1st
The difference between the third embodiment shown in FIG. 3 and the first embodiment shown in FIG. 9 is that the halogen lamps 21 m and 21 b
Light emitting diodes 35m and 35b are installed on the river,
Filter 22a.

22b and shutters 24a and 24b are not provided, and instead of a TV camera, a linear sensor and a motor control circuit 3 that controls the motor for moving the table 39 are used.
7 is provided.

The light emitting diodes 35a, 35bt; Il are turned on alternately, the pattern 3 is alternately illuminated, and the state is converted into an electrical signal by an imaging device consisting of an imaging lens 5 and a linear sensor. Since the image detection by the linear sensor is one-dimensional, the motor control circuit 37 drives the motor in accordance with the command from the synchronous control circuit, and the table 39 is moved in the longitudinal direction of the detection area of the linear sensor. It moves in a perpendicular direction (indicated by arrow C in FIG. 13) to obtain a two-dimensional image signal. At this time, the moving speed of the table 39 is determined by the detection resolution in the moving direction (arrow C) and the detection time for that line by the linear sensor.

In this way, the detected two systems of image signals are processed in the same manner as in the first embodiment shown in FIG. 9, and bright spot noise is eliminated. Regarding the processing of the two systems of image signals, the difference from the first embodiment is that when the TV left camera was used, image signals for one screen were stored in the memory 29, whereas when the linear sensor 36 was used, the image signals for one screen were stored in the memory 29. In this case, since the lighting of the light emitting diodes 35m and 35b is switched every time one line is scanned, image signals for one line are stored in the memory 29. Therefore, the memory 29 needs only a small capacity memory.

FIG. 14 is a time chart showing the operation of the third embodiment shown in FIG. 13. As shown in the figure, switching on and off of the light emitting diodes 35a and 35b, writing to and reading from the memory 4, and movement of the detection resolution unit of the table 39 are performed accurately by the power of the synchronous control circuit.

FIG. 15 is a diagram showing a fourth embodiment of the present invention.
The same parts as those in the second embodiment and the third embodiment are given the same reference numerals, and the explanation thereof will be omitted. That is, the illumination system is exactly the same as in the third embodiment, and the detection system is exactly the same as in the second embodiment except that the image signal for one line is processed as one unit.

The light emitting diodes 35m, 3 of this fourth embodiment
The switching of the light on 5b, the reading-out of the memory 4, the movement of the table 39, etc. are completely the same as in the third embodiment shown in FIG. Further, as for the image signals outputted from the linear sensor A, two systems of image signals are alternately outputted for one line each, similar to the third embodiment shown in FIG. Therefore, the image signal processing in the fourth embodiment is performed in the eleventh embodiment.
Unlike the case of the second embodiment shown in the figure, processing is not performed in units of sea screens, but in units of l lines. Therefore, memory 4 may have a small capacity.

FIG. 16 is a diagram showing a fifth embodiment of the present invention.

As shown in FIG. 16, this fifth embodiment is provided with two systems, an illumination system and a detection system. i.e., sea)
The pattern 3 on the pattern 19 is illuminated from two directions with a difference J' by two illumination systems consisting of halogen lamps 21m, 21b, filters 22m, 22b, and condenser lenses 23a, 23b. Interference filters are used as the filters 22&, 22b, and are illuminated simultaneously with monochromatic light of different wavelengths from two directions. The illuminated pattern 3 is wavelength-separated by a dichroic mirror 40 and 7' filters 41a and 41b. That is, the dichroic mirror 40 passes the monochromatic light emitted from the condenser lens Z31L and reflects the monochromatic light emitted from the condenser lens 23b. And filter 41
The filter 41a is composed of an interference filter having the same wavelength as the filter 22a, and the filter 41b is composed of an interference filter having the same wavelength as the U filter 22b. In this way, the two wavelength-separated monochromatic lights are each transmitted through the imaging lens 25''a.
, 25b to the linear sensors 36m, 36b, and are imaged on the linear sensors 36m, 36b.

The imaging magnification of the imaging lenses 25a and 25b is determined by the ratio of the required detection resolution and the size of one detection pixel of the near sensors 35m, 36b.

As a result, in the first embodiment, the detection of the pattern 3 from different directions is performed simultaneously using monochromatic light irradiated simultaneously from two different directions. At this time, since the image detection by the linear sensors 36 a and 36 b is one-dimensional, the motor control circuit 37 drives the motor A to move the table 39 in the direction of the arrow C. Performs dimensional image detection. The operation of this two-dimensional image detection is the same as in the third embodiment shown in FIG. 13 and the fourth embodiment shown in FIG. 15.

Two systems of image signals L1 output from the linear sensors 36a and 36b are each provided with a shading correction circuit 27a.
, 27 b, uneven illuminance and linear sensors 36L, 36
After removing the sensitivity unevenness of b, the binarization circuit 28 &.

It is binarized at 28b. The binarized image signals are input to an AND circuit 31 and synthesized.

The processing of the image signal in this fifth embodiment is completely the same as that in the first embodiment shown in FIG. 9 except for the following points. That is, in the embodiments w and l, an image signal for one screen obtained by irradiation from one direction (halogen lamp 21 &) is read into the memory 4, and the image signal obtained by irradiation from the other direction (halogen lamp 21&) is read into the memory 4.
The memory 4 is read out in synchronization with the image signal obtained by the halogen lamp 21b), and the two image signals are combined, but in the fifth embodiment, monochromatic light of different wavelengths is emitted from two different directions. The illumination (halogen lamp 2
11.21b) and reading the image signal into memory are not required, and a binary image free of bright spot noise can be detected at high speed in real time.

FIG. 17 is a diagram showing a sixth embodiment of the present invention.

The illumination system in the sixth embodiment is exactly the same as the fifth embodiment shown in FIG. 16, and monochromatic light of different wavelengths is irradiated onto the pattern 3 from two different directions.

The configuration of the detection system for detecting illumination by each monochromatic light up to the shading correction circuits 2Ta and 27b is the same as in the fifth embodiment. Therefore, in this sixth embodiment as well, two systems of image signals are simultaneously transmitted to the linear sensor 36m,
36b, and the shading correction circuit 27m
, 27 b with uneven illumination and linear sensor 35m, 3
The sensitivity unevenness of 5b is removed. The subsequent processing is the first
Similarly to the second embodiment shown in FIG. 1, the signal is inputted to the binarization circuit 4 via the minimum value circuit 32.

Regarding image signal processing, the difference between the sixth embodiment and the second embodiment is that the sixth embodiment performs real-time processing like the fifth embodiment, so the memory 29 is not required. That's true. Therefore, according to the sixth embodiment, a binary image free of bright spot noise can be obtained quickly in real time. Incidentally, if the output of the minimum value circuit 32 is used as is, it is possible to obtain a multi-level gradation image similarly to the second embodiment shown in FIG.

In the first to sixth embodiments described above, a TV camera or a near sensor is used as a means for forming an image signal, but the present invention is not limited to this, and for example, other known image forming methods may be used. Signal forming means may also be used.

Also, the above 3. 4th. Fifth. Regarding each of the sixth embodiments,
As in the first and second embodiments, if a pattern of bright metal particles is formed on the dark sheet 19, it is of course possible to change the pattern detection device to remove dark spot noise. It is.

〔Effect of the invention〕

As is clear from the above explanation, according to the present invention, the shape of a pattern formed by a collection of fine particles such as metal fine particles having a specular reflection surface of light can be reduced from the influence of dark spot noise and bright spot noise. It can be detected as an image that does not exist. Therefore, there is an effect that the reliability of pattern detection is greatly improved.

Furthermore, since the pattern detection device of the present invention basically uses bright field illumination, it can detect a bright optical image and can perform high-speed pattern detection.

[Brief explanation of drawings]

Figure 1 is a diagram showing an example of a conventional pattern detection device using bright field illumination, and Figures 2 and 3 are detection images when a pattern is formed of fine particles such as metal particles having a specular reflection surface. A diagram showing an example, FIG. 4 (-) is a cross-sectional view of a pattern formed by an aggregation of fine metal particles, and FIG.
b) is a partial enlarged view of the cross section of the pattern shown in FIG. 4(a), FIG. 4(C) is a view showing the reflection state of illumination light in the enlarged part of the pattern shown in FIG. 4(b), The figure shows an example of a conventional pattern detection device that uses bright field illumination.
FIG. 6 is a perspective view showing a part of a pattern with five steps;
Fig. 7 is a diagram showing an example of a conventional pattern detection device using transmitted illumination, Fig. 8 (a) + (b) + (C)
+ (d) is an explanatory diagram showing the principle of the present invention, FIG. 9 is a diagram showing the first embodiment of the present invention, and FIG. 10 (a) l (
b), (e) are time charts showing the operation of the first embodiment shown in Fig. 9, Fig. 11 is a diagram showing the second embodiment of the present invention, and Fig. 12 (11)-1 (b). ). (c) is a time chart showing the operation of the second embodiment shown in Fig. 11, @13 is a diagram showing the third embodiment of the present invention, and Fig. 14 is a time chart showing the operation of the second embodiment shown in Fig. 13. 15 is a diagram showing the fourth embodiment of the present invention. FIG. 16 is a diagram showing the fifth embodiment of the present invention.
FIG. 7 is a diagram showing a sixth embodiment of the present invention. 3... Pattern, 5... Background, 6... Bright spot, 7...
...Dark point, 8...Fly particle, 9...Power opening plane, 1
9...Sheet, addition...Light source power supply, 21a, 21b.
...Halogen lamp, 22" 122 b...Filter, 23a123b...Co/tensor lens, 24
a, 24 b--sha7ri, 25, 25 a
, 25 b-Imaging lens, ah...TV left camera 27
a, 27b... shading correction circuit, 28a,
28b, 28-binarization circuit, 29... memory, addition/
...Synchronous control circuit, 31...AND circuit, 32...
Minimum value circuit, 35 a, 35 b...Light emitting diode, 36.36a, 36b...Linear sensor, 37...
- Motor control circuit, 38... motor, 39... table. Agent Patent Attorney Tadashi Akimoto f;:1 Figure 21Y1 Figure 3 Figure 4 Figure 10 (a) Figure 10 (b) t (C) Figure 11 Figure 12 (E)) Figure 12 (C) Figure 13 Figure 14 Figure 15 v, Figure 16

Claims (1)

[Scope of Claims] 1. An illumination device that illuminates a pattern to be detected on an object from two different directions, and an optical image of the pattern to be detected illuminated from the two directions, each independently detected to generate two systems of images. 1. A pattern detection device comprising: an imaging means for converting into a signal; and a synthesizing means for synthesizing corresponding portions of a detected pattern of two systems of image signals output from the imaging means. 2. The synthesizing means includes a binarization circuit that converts two systems of image signals into binarized signals, and a binarization circuit that converts two systems of image signals into binarized signals, and
2. The pattern detection device according to claim 1, further comprising an AND circuit that performs a logical product of binary signals of the system. 3. The synthesis means comprises 2. It is characterized by being composed of a binarization circuit that converts the system image signal into a binary signal, and an OR circuit that takes the sum of the two systems of binary signals output from the binarization circuit. A pattern detection device according to claim 1. 4. The combining means is configured to include a minimum value circuit that compares the magnitude of the two systems of image signals and outputs the smaller image signal. pattern detection device. 5. The combining means is configured to include a maximum value circuit that compares the magnitudes of the two systems of image signals and outputs the entire image signal of the larger one. Pattern detection device.