JPH03111707A - Detecting method of shape of object - Google Patents

Abstract

PURPOSE:To enable improvement of precision in measurement of a shape by disposing a polarization control means and an analyzer sequentially on the optical axis of a reflected light from the surface of an object and by intercepting selectively a light being different in the number of times of reflection. CONSTITUTION:A linearly-polarized parallel beam generated by a laser light source unit 31 is passed through a lambda/2 wave plate 32 and turned into a light having an arbitrary azimuth angle and falls on a galvanomirror 34 which is made to vibrate by the drive of a motor 33, and thereby scanning of the beam is conducted. The beam reflected by the mirror 34 is condensed by a scanning lens 35 and applied to an object 36. The mirror 34 and a stage 37 are controlled synchronously and independently by a computer control so that an image of the object 36 be sensed. An image sensing system is constructed by arranging an imaging lens 38, a lambda/4 wave plate 39, an analyzer 40 and PSD 41, a position detecting means, sequentially on the optical axis of a reflected light. The lambda/4 wave plate 39 is equivalent to a polarization control means which controls the direction of polarization of the reflected light. Even when a light reflected in a plurality of times enters the lens 38, only a light reflected one time is detected by the PSD 41, according to this constitution, and thus the precision in measurement of a shape can be improved.

Description

[Detailed Description of the Invention] [Summary] Regarding an object shape detection method for measuring the shape of an object in a non-contact manner, an object that can improve the accuracy of shape measurement by detecting only one reflected light even when multiple reflections occur. The purpose of this method is to provide a shape detection method.The light generated by a light source installed at a predetermined angle is focused and irradiated onto an object, and the reflected light from the surface of the object is imaged and received. An object shape detection method comprising position detection means for outputting an electrical signal and measuring the shape of the object based on the output of the position detection means, the method comprising: linearly deflecting the light from the light source with a predetermined deflection angle; ,
A deflection control means for controlling a polarization state and an analyzer are sequentially arranged on the optical axis of the reflected light from the object surface, and the shape of the object is measured by selectively detecting light having different numbers of reflection points. Configure.

[Industrial application field]

The present invention relates to an object shape inspection device, and particularly to an object shape detection method for measuring the shape of an object in a non-contact manner.

BACKGROUND ART In recent years, methods of measuring the shape and position of an object in a non-contact manner by irradiating the object with light and detecting the reflected light have been widely used. Among these, there is a need for a photodetection method that measures the three-dimensional shape of an object, but a photodetection method that is satisfactory for a wide variety of objects has not yet been developed. If such a light detection method is developed, it can be applied to, for example, shape measurement of surface-mounted components on which a wide variety of objects are mounted.

That is, surface mount components (SMD: 5urfa
ce M.

Unted Devices are soldered while placed on a board, so if there is a slight floating bridge or unattached solder, this is likely to cause a failure, and reliable mounting inspection is desired. Such mounting inspections were generally performed visually. However, since human error is inevitable in visual inspection, there is a need for an apparatus that can automatically perform mounting inspection.

[Conventional technology]

Conventionally, two-dimensional measurement has been performed to measure the outer shape and position of an object. In this measurement method, light is irradiated from a predetermined direction, the reflected light is detected by a photodetector installed vertically above, a grayscale image is collected, and the shape is measured by image processing.

In addition, methods to measure three-dimensional shapes have been tried in the past.
There are multiple methods for detecting three-dimensional shapes.

However, when scanning light to measure three-dimensional shapes without contact, if reflected light with different number of reflections enters the light detection system, it will not be possible to detect only the correct reflected light, making it difficult to measure the exact position of the object. I can't. Although this phenomenon did not pose much of a problem when measuring the two-dimensional outline of an object, it became a very serious problem when detecting light in three-dimensional measurement.

Among the conventional three-dimensional shape measurement methods, the detection method that irradiates the object with spot light or slit light and measures the height using the principle of triangulation is a method that measures the height of an object from several liters to several tens of crn by several tens of micrometers. It is suitable as a method for high-speed, high-precision measurement with a resolution of .

As a conventional object shape detection method using this method, there is a method using a laser scanning optical system as shown in FIG. 14, for example. In the figure, a collimated laser beam Pa from a laser light source 1 is incident on a galvano mirror 3 that vibrates due to the drive of a motor 2, and after being scanned by this galvano mirror 3 as a reflected laser beam pb, a scanning lens 4
(for example, an fθ lens) and stage 5 (in the figure,
movable in the direction of the arrow)) is irradiated onto the object 6 above. Then, the reflected light Pc from the object 6 is condensed by an imaging lens 7 to a semiconductor device detection element (hereinafter referred to as P-3D) 8.
The reflected light PC is detected by the PSD 8, and the height and shading (brightness) of the object 6 are measured simultaneously.

Here, the PSD is a semiconductor element that changes the current values of two output currents I, I2 according to the position of the laser beam spot irradiated on its light receiving surface.When light is irradiated to the center of the light receiving surface, and I2 have approximately the same size, and when light is irradiated on one of both sides, the electric current is drawn from the off-centered side, and the current value increases.

That is, in the case of the example shown in FIG. 14, as the laser beam scans, light trajectories x, y, z are formed on the light receiving surface of the PSD 8.
is drawn, and this trajectory shows that Y is on the top surface of the object 6, and
, Z correspond to, for example, the upper surface of the stage 5 on which the object 6 is placed. I, , I2 output along these x, y, and z trajectories are processed according to the following equation (■■) to obtain the values of height h and brightness P, respectively.

P=1. +1. ......■ h determined by the above formula (■) represents the height of the object 6. For example, by comparing this h with a predetermined inspection reference value, the height of the object 6 can be automatically inspected.

[Problem to be solved by the invention]

However, with such conventional object shape detection methods, it is possible to perform inspection by simultaneously collecting height and shading information of the object using a laser scanning optical system at high speed. When this occurs, countermeasures are not taken into consideration, and there is a problem in that accurate height measurement may not always be possible.

That is, the object 6 includes metal parts, glass parts, etc., and multiple reflections easily occur on a reflective surface under predetermined conditions. For example, as shown in FIG. 15, when the object 6 is a surface-mounted component, the light pb incident vertically on the electrode portion 6a is reflected under certain conditions, in addition to the once-reflected light PC, the multiple reflected light Pct. occurs. Note that A in the figure represents a single reflection point, and B and B' represent twice reflection points. These lights are PSD
When the image is formed on eight planes, errors occur in both the height signal and the brightness signal from the PSD 8, making it impossible to accurately measure the height.

FIG. 16 shows a side view when multiple reflections occur.

The laser beam pb incident vertically from above is reflected at point A, and a portion of it advances in the detection angle direction and forms an image at a dot on the PSD 8 surface. Also, the reflected light that is specularly reflected at point A is specularly reflected at point B and P
An image is formed at point E on the SD8 surface. In this case, the reflected light at point B is optically the same as regular reflection from the virtual image position at point C, resulting in an error in height measurement. That is, the reflected light from point A is P
The image is formed as a point in SD8, and the reflected light from point C (point B
) is imaged at point E in PSD8, so PS
At D8, it is detected as reflected light from point G, which is an intermediate position between the two, and the light is imaged at point F, and in this case, an error occurs in height measurement.

In any case, when the PSD 8 is simultaneously irradiated with light including multiple reflections at two or more points, it generates a current corresponding to the average position of the two points, and the true height cannot be measured.

Therefore, an object of the present invention is to provide an object shape detection method that can improve the accuracy of shape measurement by detecting only one reflected light even when multiple reflections occur.

[Means to solve the problem]

In order to achieve the above object, the object shape detection method according to the present invention has the following features:
comprising a position detection means for converging light generated by a light source installed at a predetermined angle and irradiating it onto an object, forming an image of reflected light from the surface of the object, receiving the light, and outputting an electric signal according to the light receiving position, In the object shape detection method for measuring the shape of the object based on the output of the position detection means, the light from the light source is configured to be linearly deflected at a predetermined deflection angle, and
A deflection control means for controlling a polarization state and an analyzer are sequentially arranged on the optical axis of the reflected light from the object surface, and the shape of the object is measured by selectively detecting the light having a different number of reflections. I have to.

[Effect]

In the present invention, the polarization control means and the analyzer are sequentially arranged on the optical axis of the reflected light from the object surface, and the shape of the object is measured by selectively blocking the light having different number of reflections.

Therefore, even if multiple reflections occur, only one reflected light is detected, improving the accuracy of shape measurement.

[Explanation of principle]

First, the present invention will be explained in detail.

As a result of intensive studies to solve the above problems, the inventors of the present invention have found that if a method that applies changes in the polarization state that occur when linearly polarized laser light is reflected, light that has been reflected multiple times can enter the PSD. However, we have reached the fact that unnecessary reflected light of a predetermined polarization can be selectively blocked, and thereby only directly reflected light can be selected to eliminate errors in shape measurement.

Next, to explain the above principle in detail, first, Fig. 2 (a
) shows the change in polarization of the reflected light 14 when a linearly polarized laser 13 is incident on the metal surface 11, and FIG. The conduction medium metal and the transparent dielectric glass are expressed by the same formula, and the amplitude of the reflected light is R, = tan (φ-x)/1an (φ+X) ・AP ・・・・・・■R, =-sin
It is expressed as (φ-x)/5in (φ+X) ・A,...■. Here, R,,R, is the vertical direction of the reflected light,
horizontal component, φ is the angle of incidence, X is the angle of refraction, AP.

A indicates the vertical and horizontal components of the incident light, respectively.

Further, for metals, X is a complex number, and for glass, it is a real number. Therefore, in metals, RP/APR3/A3
is also a complex number, and a change in phase occurs due to reflection. As shown in FIGS. 3(a) to 3(c), even if linearly polarized light is incident on a metal, the reflected light generally becomes elliptically polarized light. Furthermore, when linearly polarized light is incident on a glass surface, the reflected light also becomes linearly polarized light.

Here, it is assumed that the incident light is linearly polarized light, and its vibration direction makes an angle α8 with the incident surface. That is, it is assumed that the following equation holds true.

tanα==As/Ap...■Also, regarding the reflected light, if α, then ta fict, =Rs
/Rp=-cos(φ-x)/cos(φ+X)・tanα1...■.

Therefore, when the incident azimuth is α8-0°, 90°,
From the above formulas 〇, ■, ■, the reflection azimuth angle α is equal to α□.

In Fig. 4 (a) to (c), the incident azimuth α with respect to the metal surface 11 is set to O', 45°, and 900, respectively, and the incident angle φ is changed to 0', 40°, 85°, and 90°. As is clear from this figure, a change occurs in the reflection azimuth α only when α=456.

On the other hand, FIG. 5 shows changes in the reflection azimuth angle on the glass surface 12, and in this case as well, the reflection azimuth angle α changes only when α1=45'.

Next, consider the case where the light is reflected twice on the metal surface 11 or the glass surface 12. Figure 6 shows a twice-reflected optical path, and when the incident azimuth α, = O'', is 90°, there is no change in polarization, as in the case of single reflection, and the polarization of the incident light is preserved. ,
The polarization direction of the light incident at α1=45' changes depending on the incident angle and the material of the reflecting surface. Therefore, the amplitude of the incident light is AP,
As, the incident angle is φ.

Then, the light reflected once on the reflecting surface H' is R□+R3l9
Let the reflection angle be φ1. Further, this reflected light is incident on the reflective surface I at an incident angle φ2, and the twice reflected light is reflected with amplitudes of R,2 and R52. Combinations of the reflective surfaces H and I are shown in Table 1 below.

Combinations of Reflective Surfaces (Table 1) Table 2 below shows the polarization state of the twice reflected light when the incident polarization is 45°.

(Hereinafter, in the margin of this page) Polarization of twice-reflected light when the incident polarization is 45° (Table 2) From this table, it can be seen that the glass-glass combination becomes linearly polarized light even if it is reflected twice, but for other cases, it becomes elliptical. It turns out that the light is polarized.

FIGS. 7 and 8 show changes in the polarization of specularly reflected light a, scattered light S, and c when incident light is reflected by the reflecting surface H. The arrows of scattered light (i) and (c) in each figure indicate the direction and intensity of scattered light within a plane including incident light and reflected light. The polarization direction of the reflected light beams S and C is the same as that of the specularly reflected light A(7) even on metal and glass surfaces.
) indicates the same direction as the polarization direction.

Based on the principles described above, the polarization separation method is applied to objects to be measured made of metal and glass. Based on the above logic, it is possible to separate the reflected lights when a plurality of reflected lights occur when one incident light enters the object. For this purpose, when the incident polarization azimuth is O'' or 90', there is no difference in polarization direction between the once reflected light and the multiple reflected light.
The angle of incidence must be between 90°.

FIG. 9 shows cases of light reflection. Regarding the reflective surface H, as shown in Table 1 above, a combination of metal and glass can be considered. Therefore, here, all the reflective surfaces I4 are metal surfaces. As shown in Figure 9,
Consider three models (a), (b), and (c). Here, (a).

Regarding (b), since the light travels in the detection direction only once reflected, there is no need to separate the lights.

However, when there is a cylindrical object 21 or an edged object 22 as shown in (c), multiple reflections occur. In this case, both the once reflected light and the twice reflected light are elliptically polarized light.

However, if there is a difference in the azimuth angle of the reflected polarized light, the two can be separated.

FIG. 1 shows a polarization separation optical system for separating both. This optical system uses two reflected lights with different polarization directions.
It is possible to allow one of them to pass and block the other, thereby attempting to block multiple reflected light. First, the two reflected lights 23 and 24 are λ/4
The light passes through a wavelength plate (corresponding to polarization control means) 25. Here, λ
The long axis direction of the /4 wavelength plate 25 and the long axis direction of the reflected light 24 to be blocked are made to match. Of the two reflected lights 23 and 24 that have passed through this wavelength plate, reflected light 24 is converted into linearly polarized light, and reflected light 23 passes through as elliptically polarized light. Then, the reflected light 24, which has become linearly polarized light, is analyzed by the analyzer 26.
The reflected light 23 of elliptically polarized light is allowed to pass through.

According to this method, it is possible to separate a plurality of elliptically polarized lights having different azimuthal angles and select only the desired reflected light. Therefore, by making the reflected light 23 correspond to the once reflected light and the reflected light 24 corresponding to the multiple reflected light, only the once reflected light is allowed to pass, eliminating the influence of the multiple reflected light, and improving the accuracy of shape measurement. can be done.

〔Example〕

Next, embodiments of the present invention based on the above basic principle will be described.

FIG. 10 is a block diagram of a first embodiment of the present invention, and shows an object shape inspection device. In the figure, a laser light source unit 31 (corresponding to a light source) is composed of a semiconductor laser and a collimating lens, and generates a parallel beam of linearly polarized light. The linearly polarized parallel beam generated here passes through the λ/2 wavelength plate 32 and is converted into light having an arbitrary azimuth angle.

Next, the galvanometer mirror vibrates due to the drive of the motor 33.
34, and the beam is scanned. Galvano mirror 3
The beam reflected by 4 is condensed by a scanning lens 35 and irradiated onto an object 36 from vertically upward (parallel to the Z axis). The object 36 is placed on a stage 37, and the galvanometer mirror 34 and the stage 37 are synchronized by computer control and controlled independently to image the object 36.

The imaging system is installed at a predetermined angle with respect to the stage 37 surface,
On the optical axis of the reflected light, an imaging lens 38, a λ/4 wavelength plate 39,
Analyzer 40 and PSD (corresponding to position detection means) 41
Configure by arranging them sequentially. The λ/4 wavelength plate 39 corresponds to polarization control means for controlling the polarization direction of reflected light.

In the above configuration, in this method, the beam is scanned one-dimensionally to simultaneously measure height and brightness.

In addition, the stage control circuit allows the stage 37
By moving in a direction perpendicular to the optical scanning direction, the three-dimensional shape of the object 36 can be measured.

In this measurement, if multiple reflections occur due to light reflected from the edges of the object 36, for example, the polarization separation method as described in the basic principle section above is applied, and the imaging lens 36
Even if light that has been reflected multiple times enters the PSD 4, by using the λ/4 wavelength plate 39 and the analyzer 40, it is possible to selectively block the light corresponding to multiple reflections of a predetermined polarized light, and only the once reflected light can be transmitted to the PSD4.
1, the accuracy of shape measurement can be improved. Furthermore, since rapid processing is performed using the polarization separation method to avoid the influence of multiple reflections, the height and brightness of the object 36 can be detected at high speed (0.5M/pixel).

Figure 11 shows a second embodiment of the present invention, in which the galvanometer mirror is replaced with a rotating mirror 45. That is, the parallel beam that has passed through the λ/2 wavelength plate 32 is scanned at high speed by the side surface of the rotating mirror 45, is reflected by the reflection mirror 46, passes through the fθ lens 47, is further reflected by the reflection mirror 48, and is directed toward the object 36. is irradiated. This embodiment has the advantage that by rotating the rotary mirror 45, the object 36 can be scanned at a higher speed than in the first embodiment.

Figure 12 shows a third embodiment of the present invention, in which a slit light is projected instead of a spot light. That is, after the light from the laser beam unit 31 is reflected by the reflection mirror 49, it is polarized through the λ/2 wavelength plate 50 and enters the cylindrical lens 51. Then, the spot light is converted into an elongated slit light by the cylindrical lens 51, and the light is directly projected onto the object 36.

The reflected light of the light projected onto the object 36 is reflected by an imaging lens 38,
The light passes sequentially through a λ/4 wavelength plate 39 and an analyzer 40, and is received by a COD camera (corresponding to position detection means) 52.

Therefore, in this embodiment, since it is not a spot, scanning of light is unnecessary.

Figure 13 is a diagram showing a fourth embodiment of the present invention, and this embodiment performs surface irradiation. That is, the light from the laser beam unit 31 is expanded in the plane direction by the beam expander 53 and enters the λ/2 wavelength plate 32, and then,
The target object 36 is illuminated with a surface. The surface-irradiated light is received on a surface-by-surface basis by a CCD camera 52 with a configuration similar to that of the third embodiment. Therefore, this embodiment has the advantage that two-dimensional scanning can be performed instantaneously.

Note that in each of the above embodiments, light is irradiated onto the object from a vertical direction and light is detected from an oblique direction, but the invention is not limited to this, for example, light may be irradiated from an oblique direction and light is detected from vertically above. Alternatively, the light may be detected from an oblique direction. In this case, the incident angle and detection angle of light may be set arbitrarily.

In addition, the optical position detector may be installed at predetermined angles in multiple directions relative to the target object to detect light. In this way, the blind spot, which is a drawback of so-called triangulation, will be eliminated and shape measurement will be easier. Improves accuracy.

Furthermore, a mechanism may be provided to rotate the incident light source wavelength plate, the light detection wavelength plate, and the analyzer with respect to the optical axis, so that they can be adjusted to optimal angles.

Further, as the polarization control means, for example, a light modulation element may be used instead of a wavelength plate, and the polarization direction may be electrically controlled to a predetermined angle.

〔Effect of the invention〕

According to the present invention, even if multiple reflections occur, it is possible to detect only one reflected light to eliminate errors caused by multiple reflections, and it is possible to improve the accuracy of measuring the shape of an object.

[Brief explanation of drawings]

Figures 1 to 9 are diagrams explaining the present invention in detail. Figure 1 is a diagram showing its polarization separation optical system, Figure 2 is a diagram showing the state of its reflected light, and Figure 3 is its reflected light. Figure 4 shows the polarization change due to reflection from a metal surface, Figure 5 shows the polarization change due to reflection from a glass surface, and Figure 6 shows the optical path of the twice reflected light. Figure 7 is a diagram explaining the polarization of the scattered light, Figure 8 is a diagram explaining the polarization of other scattered lights, Figure 9 is a diagram explaining the reflection model of the light, and Figure 10 is a diagram explaining the polarization of the scattered light. FIG. 11 is a block diagram of the first embodiment of the object shape detection method according to the present invention. FIG. 11 is a block diagram of the second embodiment of the object shape detection method according to the present invention. FIG. 13 is a configuration diagram of a fourth embodiment of the object shape detection method according to the present invention. FIGS. 14 to 16 are diagrams showing a conventional object shape detection method. is its overall configuration diagram, Figure 15 is a diagram explaining the reflected light of multiple reflections, and Figure 16 is a diagram explaining the reflected light of multiple reflections.
The figure is a diagram illustrating the path of multiple reflections. 23.24... Reflected light, 25.39... λ/4 wavelength plate (polarization control means)
26.40... Analyzer, 31... Laser beam unit (light source), 32.
50... λ/2 wavelength plate, 33... Motor, 34... Galvano mirror 35... Scanning lens, 36... Target object , 37...stage, 38...imaging lens, 41...PSD (position detection means), 45...
...Rotating mirror 46.48.49...Reflecting mirror 47...
...fθ lens, 51...Cylindrical lens, 52...
... CCD camera (position detecting means) 53 ... Beam expander - Diagram showing the situation of reflected light for detailed explanation of the present invention Fig. 2 Reflected light for detailed explanation of the present invention Fig. 3 shows the polarization change of the reflected light; Fig. 3 shows the twice-reflected optical path for the detailed explanation of the present invention. Diagram showing the polarization change of φ=90°a; Regularly reflected light b: Scattered light C: Scattered light FIG. 7 illustrating the polarization of scattered light for detailed explanation of the present invention. Fig. 0 A block diagram of the third embodiment Fig. 2 A block diagram of the fourth embodiment Fig. 14 Fig. d Fig. 5 explaining the reflected light of multiple reflections in the conventional example

Claims (1)

[Claims] Light generated by a light source (31) installed at a predetermined angle is focused and irradiated onto an object (36), and the reflected light from the surface of the object (36) is imaged and received. Position detection means (41) that outputs an electrical signal according to the position
, 52), and detects the object () based on the output of the position detection means (41, 52).
36) In the object shape detection method for measuring the shape of the object (36), the light from the light source (31) is configured to be linearly deflected with a predetermined deflection angle, and the light reflected from the surface of the object (36) is deflected on the optical axis. , a deflection control means (39) for controlling the deflection state, and an analyzer (40)
A method for detecting the shape of an object, characterized in that the shape of the object (36) is measured by sequentially arranging the light beams and the light beams having a different number of reflections.