JP2643744B2 - Image position detection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image position detecting device applicable to a device used in the manufacturing process of a semiconductor device or a liquid crystal display device, for example, a film thickness measuring device, etc. The present invention relates to an apparatus for enlarging an image of a formed predetermined pattern with an optical microscope, further capturing the enlarged image as an input image, and matching the input image with a previously registered reference image to detect the image position. .

[0002]

2. Description of the Related Art In a technical field such as analysis and measurement, when a measurement is performed while aiming at a desired portion of an object to be measured (hereinafter referred to as a "measurement site"), a measuring device measures the portion to be measured. In general, a pattern matching method is employed for positioning at a position where the measurement is performed (hereinafter, referred to as a “measurement execution position”). This is because an image of an area on the object to be measured having a predetermined positional relationship with the part to be measured (hereinafter, referred to as a “reference image”) is registered in advance, and an image of the object to be measured is picked up from among images obtained. The position matching the reference image is detected, and the position of the measurement site is specified.

[0003]

However, when position detection is performed on an object on which a transparent thin film is formed, for example, as in a film thickness measuring device for measuring the thickness of a transparent thin film formed on a semiconductor wafer. According to the conventional pattern matching method, position detection based on collation between an input image and a reference image becomes extremely unstable, and reliability is reduced.
Specifically, when repeatedly measuring the line width and the thickness of an insulating film of a predetermined portion of a circuit pattern for each of a plurality of chips formed side by side on a semiconductor wafer or for each semiconductor wafer, a conventional method is used. In this case, a portion to be detected as corresponding to the reference image was not detected, and the measurement of the measured portion to be measured was not performed in some cases.

[0004] Then, as a result of earnest research by the present inventors, the following has been found. That is, with respect to the DUT on which the transparent thin film is formed, even if the same pattern is formed, the obtained input image is at least partially shaded (that is, the intensity of light) of the pattern. May be reversed. In such an input image in which the shading relationship is reversed, even if the pattern is compared with the same reference image, the two are not considered to be coincident with each other, and it is not detected that the input image corresponds to the reference image. Was. It has been found that the inversion of the shading relationship is mainly caused by unevenness in the thickness of the transparent thin film formed on the object to be measured. Hereinafter, the reason will be described with reference to FIGS.

FIG. 13 is a view showing an example of a semiconductor wafer to be measured. In this semiconductor wafer, as shown in FIG. 1, a silicon nitride film 1a and a silicon oxide film layer 2 which are divided into a predetermined shape and patterned are formed on a semiconductor wafer 1. For example, as shown in FIG. 14A, an image of a portion indicated by a two-dot chain line in FIG. 13A is an image region (region 1) corresponding to a part of the silicon nitride film 1a and a silicon oxide film layer 2. And an image area (area 2) corresponding to a part of. Here, when the light intensity of the input image is examined along the alternate long and short dash line in FIG. 14A, as shown in FIG. 14B, the intensity of the region 1 is larger than that of the region 2. . However, when the film thickness becomes uneven and the film thickness t of the silicon oxide film layer 2 (see FIG. 13B) changes, FIG.
As shown in FIG. 4 (c), the intensity of the reflected light from the area 2 is increased and the light and shade relation with the area 1 is reversed. As a result, the light and shade relation between the two areas is reduced as compared with the case of FIG. 14 (b). May be reversed.

By the way, when considering a certain wavelength, it can be easily understood that the density inversion occurs due to the change in the film thickness as described above. However, in general image position detection, white light is used as a light source, so that white light is used as a light source. There is a view that image density inversion may not occur. This view indicates that, as shown in FIG. 15, the intensity of the input image of the DUT on which the transparent thin film is formed is an integrated value of each wavelength component (a value corresponding to the area of the region shown by oblique lines in FIG. 15). Even if the film thickness t is different, it is based on the theory that the intensity of the input image becomes substantially equal and the density inversion of the image does not occur.

However, in the image position detecting device, an image is picked up through an image pickup means such as a camera, so that the input image is affected by the wavelength characteristic of the camera itself and the change in the film thickness t is obtained. The resulting influence is multiplied by the wavelength characteristics of the camera. Therefore, even when white light is used as a light source, the above-described grayscale inversion is often observed in an actually obtained input image.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and provides an image position detecting apparatus and method capable of accurately detecting the position of an object to be measured on which a transparent thin film is formed. Aim.

[0009]

According to a first aspect of the present invention, there is provided an illumination optical system for irradiating a predetermined illumination position with illumination light, and a transparent thin film formed on one principal surface thereof. A stage that positions an arbitrary imaging region of the measurement object at the illumination position while mounting the measurement object, and forms an image of the imaging region by condensing reflected light from the imaging region at a predetermined position. An image forming optical system, an image pickup means for picking up an image of the image pickup area to obtain an input image, and the same pattern picked up by the image pickup means, and at least a part of the light and shade relationship is in a mutually inverted state. Storage means for storing the two images as first and second reference images, respectively;
When the partial image of the input image substantially matches at least one of the first and second reference images, the first or second one of the partial images of the input image is determined.
Means for specifying a partial image position that matches the reference image.

[0010]

According to the present invention, two images having the same pattern and in which the shading relationship is partially inverted are stored as the first and second reference images, respectively. In addition, an image of the imaging region of the device under test is captured, and an input image is obtained. Then, the position of the pattern corresponding to the reference image on the input image is detected when the partial image of the input image substantially matches at least one of the first and second reference images. Therefore, even if the grayscale reversal occurs partially in the input image, the position can be detected by collation with the first or second reference image.

[0011]

FIG. 1 is a view showing a film thickness measuring apparatus to which the image position detecting apparatus according to the present invention can be applied, and FIG. 2 is a block diagram of the film thickness measuring apparatus. In the following description, an embodiment of the image position detecting device will be clarified by describing the configuration and operation of the film thickness measuring device.

A. Configuration of Film Thickness Measurement Apparatus This film thickness measurement apparatus includes an illumination optical system 10 and an imaging optical system 20.
It has. The illumination optical system 10 is provided with a light source 11 that emits white light. The light from the light source 11 is supplied to a condenser lens 12, an aperture stop 13, and a field stop 1.
Image forming optical system 2 via the condenser lens 4 and the condenser lens 15
It is incident on zero.

The imaging optical system 20 includes an objective lens 21, a beam splitter 22, and an imaging lens 23. Illumination light from the illumination optical system 10 is reflected by the beam splitter 22, and passes through the objective lens 21 to a predetermined illumination position. IL
Is irradiated. In FIG. 1, reference numeral 24 denotes a pupil position.

An XY stage 30 is arranged near the illumination position IL. The XY stage 30 responds to a control signal from an XY stage drive circuit 31 while mounting a semiconductor wafer 1 formed by patterning thin films of a silicon nitride film 1a and a silicon oxide film 2 as an object to be measured. To move in the X and Y directions to position an arbitrary region (a region where imaging is desired) on the surface of the semiconductor wafer 1 at the illumination position IL. Although not shown in the drawings,
The position (X, Y coordinates) of the XY stage 30 is detected, and the position information is given to a control unit 60 that controls the entire apparatus.

The light reflected from the imaging region of the semiconductor wafer 1 located at the illumination position IL is condensed at a predetermined imaging position via the objective lens 21, the beam splitter 22, and the imaging lens 23, and is focused on the imaging region. Is enlarged and projected.
A reflecting mirror 40 having a pinhole 41 at the center is disposed near the image forming position. Therefore, of the reflected light, the reflected light LS that has passed through the pinhole 41 is incident on the spectral unit 50.

The spectroscopy unit 50 includes a diffraction grating 51 for dispersing the reflected light LS, and a photodetector 52 for detecting a spectrum of the diffracted light diffracted by the diffraction grating 51. As the diffraction grating 51, for example, a flat field type diffraction grating that forms a spectral spectrum on a plane, a diffraction grating with a sweeping mechanism, or the like can be used. On the other hand, the photodetector 52 is configured by, for example, a photodiode array, a CCD, or the like, and is arranged in a conjugate relationship with the pinhole 41. For this reason, the spectroscopy unit 5
The light LS captured by the light source 0 is split into a diffraction grating 51, and a signal corresponding to the spectrum of the light LS is output to the photodetector 5.
2 to the arithmetic circuit 53 (FIG. 2). In the arithmetic circuit 53, based on the signal, a conventionally well-known method (the description thereof is omitted here) is used.
The film thickness of the thin film (not shown) formed in the above is obtained, and the result is output to the control unit 60.

On the other hand, of the light from the semiconductor wafer 1, the reflected light reflected by the reflecting mirror 40 enters the image pickup unit 70. In this imaging unit 70, the reflected light from the reflecting mirror 40 is condensed on an imaging surface 72 a of a two-dimensional imaging camera 72 via a relay lens 71. Thereby, an image of the imaging region of the semiconductor wafer 1 is formed. Note that an image (input image) captured by the two-dimensional imaging camera 72
Is supplied to the image processing unit 73, and after being appropriately processed by the image processing unit 73, is displayed on the CRT 74 and output to the control unit 60.

The control unit 60 includes, as shown in FIG.
A well-known CPU 61 for executing a logical operation;
And a memory 62 for temporarily storing various data during operation of the apparatus, while preliminarily storing various programs for controlling the device 1. These CPU 61 and memory 62
Are connected to each other by a common bus 63, and are also connected to an input / output port 64 via the common bus 63. The input / output port 64 allows the control unit 60 to input / output with the outside, for example, the panel key 65, the XY stage drive circuit 31, and the arithmetic circuit 53.
And a signal is exchanged with the image processing unit 73.

B. Operation of Film Thickness Measuring Device Next, the operation of the film thickness measuring device of FIG. 1 will be described with reference to the flowchart of FIG. First, prior to the actual film thickness measurement, registration of a measurement target site where film thickness measurement is performed and registration of a reference image for image position detection are performed (Step S).
1).

FIG. 4 is a flowchart showing the procedure of the registration process. In performing this registration process, first, the XY-axis is set while positioning the semiconductor wafer 1 for registration with a certain accuracy by a handling machine (not shown).
It is transported on the stage 30 (step S11). Then, the operator operates the panel key 65 to display the CRT 7
4 is a measurement execution position of the film thickness measuring device substantially at the center of the input image,
The XY stage 30 is moved so that the part to be measured is positioned near an image (a pinhole image 41a in FIG. 5) such as a black dot-shaped shadow appearing in the input image due to the formation of 1. Thereafter, when the operator presses a registration key (not shown) provided on the panel key 65, the coordinate value of the XY stage 30 is stored in the memory 62 via the input / output port 64 as the approximate coordinate value of the measured part. Is performed (step S12).

In step S13, it is determined whether or not the number of readings of the measured portion has reached a predetermined value, and the processing in step S12 is repeatedly executed until the number of readings reaches the predetermined number. Thus, the approximate coordinate values of one or a plurality of measurement sites are temporarily stored in the memory 62. The storage of the plurality of measurement sites is performed, for example, when a predetermined measurement site is measured for each of the plurality of chips formed side by side on the semiconductor wafer 1.

When the reading of the coordinates of the last part to be measured is completed as described above, the vicinity of the part to be measured on the semiconductor wafer 1 is displayed substantially at the center of the CRT 74. More specifically, as shown in FIG. 5, the CRT 74 has an input image of the vicinity of the measured portion of the semiconductor wafer 1 imaged by the imaging unit 70, that is, an area A1 of the silicon oxide film and an image of the silicon nitride film. The area A2 is shown. Then, superimposed on the input image, a cross mark line 81,
A frame type mark line 82 surrounding a part of the image is displayed.

The marking lines 81 and 82 are electrically created by the image processing unit 73 and displayed so as to be superimposed on the image of the semiconductor wafer 1. By operating the panel keys 65, each of the marking lines 81 and 82 is independently moved within the image. Is what you do.

The operator operates the panel key 65 to
A cross-shaped mark 81 is aligned with the measured portion on the semiconductor wafer shown in the image, and a characteristic pattern region (for example, the same pattern as FIG. 14A; pattern A) near the measured portion is defined. It is surrounded by a frame type mark line 82 (FIG. 6).
In this example, the cross mark line 8 of FIG.
The case where the film thickness at the position indicated by 1 is measured is shown.

When the operator depresses the registration key, the partial image surrounded by the frame-shaped mark 82 is stored in the memory 62 as the first reference image, and the frame-shaped mark is The relative position data with respect to the cross mark 81 is stored in the memory 62 (step S14).

Next, in step S 15, for example, for a portion to be measured in another chip on the same semiconductor wafer 1, the same pattern as the above-mentioned pattern A is used except that the density inversion (this density inversion is performed on the same semiconductor wafer 1). Even in this case, there is a portion having a pattern (for example, the same pattern as in FIGS. 14A and 14C; pattern B) in which a portion of the formed thin film has uneven thickness. The image is stored in the memory 62 as a second reference image. Specifically, the operator operates the panel key 65 while looking at the CRT 74 to move the XY stage 30 so that the pattern B is projected on the CRT 74, and then moves the frame 82 to move the pattern B by the frame 82. Surround,
In addition, a cross mark 81 is placed on the portion to be measured on the semiconductor wafer 1.
And further press the key for registration.

Thereafter, the semiconductor wafer 1 for registration is carried out to the original storage position by the handling machine (step S16).

Thus, the process of storing in memory 62 the two partial images (first and second reference images) having the same pattern and in which the shading relationship is mutually inverted, and the position of the measured part, The registration process is completed.

Next, returning to FIG. 3, description of the film thickness measuring operation will be continued. When the registration process is completed as described above, the semiconductor wafer 1 to be subjected to film thickness measurement is transported onto the XY stage 30 while being positioned with a certain accuracy by the handling machine (step S2).

Then, in step S3, the approximate coordinate values of the measured part are read from the memory 62, and the semiconductor wafer 1 is positioned based on the coordinate values. As a result, a nearby area (imaging area) including the measured part is located at the illumination position I.
L (FIG. 1) is positioned with the accuracy of the approximate coordinate values, an image of the imaging area is displayed on the CRT 74,
The input image is temporarily stored in the memory 62.

Next, in step S4, the input image and the first and second reference images are read from the memory 62, and
The position of the reference image in the input image is detected according to the flowchart of FIG. Hereinafter, the procedure of position detection will be described in detail.

In step S41, a value C (H) is set as a threshold value serving as a criterion described later. In this embodiment, two types of thresholds C (H) and C (L) having different criterions are stored in the memory 62 in advance, and the stricter criterion (the threshold C ( H)), and if the position cannot be detected based on the position, then a looser criterion (threshold C (L)
) To perform position detection.

In step S42, the input image and the first reference image are read out from the memory 62, and search matching is performed based on the first reference image (registered pattern A) to obtain a correlation coefficient C1. That is, as shown in FIG. 9, over the entire area of the input image Q (X, Y), the first reference image r (x,
y) and a correlation coefficient (information indicating the degree of matching) C1 between the partial image q (x, y) of the input image Q (X, Y) are obtained by a conventionally known calculation method.

Then, in step S43, the correlation coefficient C1 obtained as described above is compared with a threshold value C (H). Here, the existence of the correlation coefficient C1 satisfying the inequality (C (H) <C1) means that the same pattern as the first reference image exists at the coordinate position. Note that, at the time of search matching over the entire area of the input image Q (X, Y) as described above, the first reference image r (x,
If a similar pattern different from y) but by chance exists in the input image, a plurality of partial images having a correlation coefficient exceeding the threshold value C (H) may appear. It is assumed that the same pattern as the first reference image exists at the coordinate position of the partial image showing the maximum correlation coefficient.

If the inequality expression (C (H) <C1) is not satisfied, that is, if the position cannot be detected using the first reference image, the process of the next step S44 is executed.

In step S44, the second reference image (registered pattern B) is read from the memory 62 instead of the first reference image, and the input image Q is read in the same manner as in step S42.
Search matching is performed over the entire area of (X, Y) based on the second reference image (registered pattern B) to obtain a correlation coefficient C2.

Then, in step S45, step S4
3, the correlation coefficient C2 obtained as described above is obtained.
Is compared with a threshold value C (H). As described above, in this embodiment, since the matching is performed using the first and second reference images having the same pattern and the grayscale relationship being in a mutually inverted state, the grayscale inversion of the input image occurs. Even in this case, the position of the input image can be detected with high probability.

However, when the image quality of the input image or the first or second reference image is poor, relatively high correlation coefficients C1 and C2 cannot be obtained by the position detection processing, and the inequality (C (H) < C1) and (C (H) <C2) may not be satisfied, and position detection using the first and second reference images may not be performed.

To deal with such a case, the following steps S46 to S52 are executed in this embodiment. That is, in step S46, the threshold value C (L) is read from the memory 62.
Is read. Then, in the same manner as in step S42, the correlation coefficient C1 based on the first reference image (registered pattern A) is obtained (step S47), and then, in step S47.
The correlation coefficient C2 is obtained based on the second reference image (registered pattern B) in the same manner as in step 44 (step S48).

Subsequently, in step S49, it is determined whether at least one of the inequalities (C (L) <C1) and (C (L) <C2) is satisfied. If step S49 is not satisfied, that is, if it is determined that position detection cannot be performed even when the threshold value is lowered, an error is displayed on the CRT 74 to notify the operator of that (step S50). .

On the other hand, if it is determined in step S49 that the above condition is satisfied, then the inequality (C (L) <C1),
It is determined whether both (C (L) <C2) are satisfied (step S51). Here, only when it is determined that both are satisfied, the process proceeds to step S52, and the coefficient C
The larger coefficient is selected from 1, C2. That is,
It is assumed that a corresponding reference image (registered pattern) exists at a position showing a coefficient of a large value, and the coordinate position is set as an image position. If it is determined that both are not satisfied, one of the equations is satisfied, and the coefficient of the satisfied equation is selected. That is, it is assumed that the corresponding reference image (registered pattern) exists at the position indicating the coefficient, and the coordinate position is set as the image position.

As described above, when an appropriate correlation coefficient is obtained, that is, when it is determined in step S43 that (C (H) <C1) is satisfied, the correlation coefficient C1,
If it is determined in step S45 that (C (H) <C2) is satisfied, the correlation coefficient C2 is determined. If it is determined in step S51 that the above condition is not satisfied, the correlation coefficient C1 is determined.
When (> C (L)) or C2 (> C (L)) or the correlation coefficient selected in step S52 is obtained, the process of step S53 is executed. In this step S53, the coordinate value corresponding to the obtained correlation coefficient is determined as the coordinates of the part corresponding to the reference image.

In this way, the position of the pattern corresponding to the reference image in the input image (the coordinates of the part corresponding to the reference image) can be detected. Then, at the same time, the coordinates of the measured part in the input image are calculated from the reference image stored in the memory 62 and the relative position data of the cross-shaped mark with respect to the reference image.

Next, returning to FIG. 3, the description of the film thickness measuring operation will be continued. When the position detection is completed as described above, the determined coordinates of the measured part and the measurement execution position MP
The distance to (FIG. 1) is calculated. Here, the “measurement execution position MP” is a position that is optically conjugate with the pinhole 41 in the film thickness measurement device, and is a preset position. Therefore, when the measurement site is positioned at the measurement execution position MP, the reflected light from the measurement site passes through the pinhole 41 and is incident on the spectroscopic unit 50, and the film thickness measurement at the measurement site is performed. It becomes possible.

Then, in step S5, based on the calculation result, the XY stage 30 is moved to position the part to be measured at the measurement execution position MP. At this time, CRT74
On the screen of, the measured portion specified in step S14 or S15 overlaps the pinhole image 41a. And then, as described above, pinhole 4
1, the reflected light LS incident on the spectroscopic unit 50 is separated, its spectral spectrum is obtained, and the film thickness at the measurement site is obtained from the spectral spectrum (step S).
6).

Next, in step S7, it is determined whether or not the film thickness measurement has been completed at all the measurement sites. If it is determined that the measurement has not been completed, the process proceeds to step S3.
The operations of steps S3 to S6 are repeated, and the film thickness is measured at each measurement site. On the other hand, when the measurement of the film thickness at all the measurement sites is completed, the process proceeds to step S8, and the semiconductor wafer 1 is carried out to a predetermined position.

Then, in step S9, it is determined whether or not the film thickness measurement has been completed for all the semiconductor wafers 1.
Steps S <b> 8 to S <b> 8 are repeated, and a predetermined film thickness measurement is continuously performed on each semiconductor wafer 1.

C. Effect of Image Position Detection Processing by Film Thickness Measuring Apparatus As described above, in this embodiment, two partial images (first and second reference images) having the same pattern and in which the shading relationship is mutually inverted. Is registered in advance, and an image of the semiconductor wafer 1 as an object to be measured is captured, and an input image is obtained. Then, the selected partial image is continuously selected over the entire area of the input image while the selected partial image is selected. Since the correlation coefficient is obtained by sequentially collating with the first and second reference images, and further based on the correlation coefficient, the position of the reference image in the input image and the coordinates of the part to be measured are specified. ,
Even if the grayscale inversion of the input image occurs, the position can be accurately detected. Therefore, in the film thickness measuring device, by performing the image position detection processing, an effect is obtained that the film thickness of a desired measured portion can be measured with high accuracy.

D. Others In the above embodiment, the case where the image position detecting device according to the present invention is applied to a reflection type film thickness measuring device that performs film thickness measurement based on reflected light has been described. In addition to this, the present invention can also be applied to a transmission type film thickness measuring device that illuminates an imaging region from the back surface side of an object to be measured and obtains a film thickness based on light transmitted through the region. It can also be applied to the device of the above.

Further, in the above-described embodiment, the case of the film thickness measuring device using the white light source 11 has been described. However, the same effect can be obtained by the film thickness measuring device using the monochromatic light source.

Further, in the above embodiment, the case where the colorless and transparent silicon oxide film is used as the thin film formed on the object to be measured is described, but a colored and transparent thin film may be used.

Further, in the above embodiment, a relatively simple pattern consisting of only two regions (region 1 and region 2) (FIG. 1)
4) is used as the first and second reference images, but has a relatively complicated pattern (for example, the pattern shown in FIG. 10) composed of three or more regions, and the shading relationship is partially The two images in the inverted state are respectively
It is also possible to perform the position detection processing as the second reference image and, in this case, satisfy the above condition (the same pattern and mutual density inversion occur).
One or more reference images are registered in advance, and the position detection processing can be performed in the same manner as described above.

When the first and second reference images are registered, in the above embodiment, the operator operates the panel key 65 while viewing the CRT 74 to register the first and second reference images. ing. On the other hand, after registering the first reference image, the image data of the reference image is
For example, it is also possible to create the second reference image by inverting the image by electrical processing in the image processing unit 73. However, it may not be preferable to use the second reference image simply created by electrical processing for the following reasons. In actual use, the first and second reference images are used as described above. Is preferably registered by the operator.

This is because if an aluminum film or an edge portion of a pattern exists in the reference image, the actual pattern differs from the electrically processed pattern.
This will be described with reference to FIGS. The shaded areas shown in FIG. 10 and FIG. 11 indicate the shading of the image projected on the CRT 74, and are shown in accordance with the density scale of FIG. 12 to indicate the shading degree.

<When an Aluminum Film is Formed> When an image of an area having an aluminum film is taken, for example, as shown in FIG.
As shown in (a), an aluminum image area 90 corresponding to the aluminum film appears in the image. Since the reflectance of the aluminum film is relatively high, it is relatively bright on the image. Also,
When an image of an area having the same pattern and where the density inversion occurs is actually taken, an image shown in FIG. Also in this image, even if the density inversion occurs between the other image areas 91 and 92, the aluminum image area 9
The 0 image remains relatively bright. On the other hand, when the density inversion processing of the image of FIG. 10A is simply performed by electrical processing, as shown in FIG. , 92 are significantly darker. As described above, the density-reversed image obtained by the electrical processing is completely different from the image actually observed.

<When an Edge is Existed> When an image of a region having an edge of a thin film is taken, the image is, for example, shown in FIG.
As shown in FIG. 1A, the edge image area 93 corresponding to the edge portion is significantly darker than the other image areas 94 and 95. The reason is that most of the illumination light is reflected in a direction different from the incident direction because the exposed surface is inclined in the edge portion, and the light returning to the objective lens 21 of the imaging optical system 20 is only a small part. Because. This means that the image area 9
The same applies to the case where the density inversion occurs at 4,95.
As shown in (b), the edge image area 93 remains dark. However, when the image of FIG. 7A is simply electrically inverted, the edge image area 9 is changed as shown in FIG.
3 becomes brighter than the other image areas 94 and 95, and becomes different from the actually observed image as in the case of the aluminum film.

Therefore, when registering the first and second reference images, it is desirable to register an appropriate image while actually viewing the image on the CRT 74 as described above.

[0058]

As described above, according to the image position detecting apparatus according to the present invention, two images having the same pattern and at least partially having the grayscale relationship mutually inverted are the first and second images. Since a partial image of the input image obtained by imaging the image of the imaging region of the device under test specifies the image position of the imaging region, the grayscale inversion of the input image occurs. Even if
Position detection can be performed with high accuracy by collation with the first or second reference image.

[Brief description of the drawings]

FIG. 1 is a diagram showing a film thickness measuring apparatus to which the image position detecting method according to the present invention can be applied.

FIG. 2 is a block diagram of the film thickness measuring device.

FIG. 3 is a flowchart showing an operation of the film thickness measuring device of FIG.

FIG. 4 is a flowchart illustrating a procedure of a registration process.

FIG. 5 is a diagram illustrating a registration procedure of first and second reference images.

FIG. 6 is a diagram illustrating a registration procedure of first and second reference images.

FIG. 7 is a flowchart illustrating a procedure of a position detection process.

FIG. 8 is a flowchart illustrating a procedure of a position detection process.

FIG. 9 is a schematic diagram showing a procedure of search matching.

FIG. 10 is a diagram showing an image of a region including an aluminum film.

FIG. 11 is a diagram showing an image of a region having an edge portion.

FIG. 12 is a schematic diagram showing the density scale of FIGS. 10 and 11;

FIG. 13 is a view showing a semiconductor wafer which is one of the objects to be measured.

FIG. 14 is a diagram illustrating density inversion of a captured image.

FIG. 15 is a graph showing a change in intensity according to a change in film thickness.

[Explanation of symbols]

 Reference Signs List 1 semiconductor wafer 10 illumination optical system 20 imaging optical system 30 XY stage 60 control unit 62 memory 70 imaging unit

Claims (1)

(57) [Claims] 1. An illumination optical system for irradiating a predetermined illumination position with illumination light, and an object to be measured having a transparent thin film formed on one of its main surfaces, and an arbitrary imaging area of the object to be measured is mounted. A stage positioned at the illumination position, an imaging optical system that focuses reflected light from the imaging region at a predetermined position to form an image of the imaging region, and captures an image of the imaging region and inputs the image. Imaging means for obtaining an image; and storage means for storing, as first and second reference images, two images of the same pattern imaged by the imaging means, and at least partially in a mutually inverting shade relationship. And that the partial image of the input image substantially matches at least one of the first and second reference images.
Means for identifying a partial image position that matches the reference image of
An image position detecting device comprising: