JP2002288661A - Image detection method and device, and wafer-processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To precisely detect a mark substantially same as a template image from an input image. SOLUTION: This image detection device 20 detects a first region approximate part which approximates a first region, including a first mark given in advance. The image detecting device 20 is provided with a first region candidate part detecting means 24 for detecting a first region candidate part which includes the image approximating to the first mark from the input image; and a second region candidate part detecting means 32, provided with the information related to the image of a second region of which a relative position to the first region has been well known, for determining the second region candidate part from the input image on the basis of the relative position, and for detecting whether the second region candidate part agrees with the information related to an image of the second region.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an image detection method, an image detection device, and a wafer processing device. In particular, the present invention relates to an image detection method, an image detection device, and a wafer processing device capable of accurately detecting a mark substantially identical to a template image from an input image.

[0002]

2. Description of the Related Art When a circuit pattern is exposed on a wafer such as a semiconductor substrate, a mark for positioning is formed in advance at a predetermined position on the wafer in order to perform accurate positioning. Is detected. And
A predetermined pattern is exposed on the wafer based on the position of the detected mark. Image matching techniques are used to detect the position of the mark on the wafer. In the conventional image matching technique, an image including a mark of a wafer is acquired as an input image, and a pixel value of the input image is two-dimensionally compared with a pixel value of a template image.

[0003]

However, usually, in order to detect the position of a mark in an input image, a cross-correlation value or the like normalized using a complicated calculation formula is obtained.
If the pixel values of the input image were two-dimensionally compared with the pixel values of the template image, an enormous number of calculations had to be performed. Therefore, it takes a lot of time to detect the position of the mark on the wafer, and it has been difficult to perform a rapid wafer exposure process.

Further, since various resists and films adhere to the wafer surface, the input image may be deformed under the influence of the reflected light on the wafer surface. In such a case, the position of the mark on the wafer can be accurately determined. Could not be detected.

Accordingly, an object of the present invention is to provide an image detecting method, an image detecting apparatus, and a wafer processing apparatus which can solve the above-mentioned problems. This object is achieved by a combination of features described in the independent claims. The dependent claims define further advantageous embodiments of the present invention.

[0006]

According to a first aspect of the present invention, there is provided an image detecting method for detecting, from an input image, a first area approximation portion approximating a first area including a predetermined first mark. A first region candidate portion detecting step of detecting a first region candidate portion including an image approximate to the first mark from the input image, and an image of a second region having a known position relative to the first region. Information is provided in advance, and based on the relative position, a second region candidate portion is determined from the input image, and it is detected whether the second region candidate portion matches information on the image of the second region. A second region candidate portion detecting step.

[0007] The second area candidate portion detecting step includes detecting whether or not the second area candidate portion matches information on the image of the second area using the image of the second mark included in the second area. Good.

The second area candidate portion detecting step uses the image of the second mark included in the second area and having a shape different from that of the first mark so that the second area candidate portion matches information on the image of the second area. Whether or not to do so may be detected.

In the image detecting method, a first area and a second area are respectively selected from a template image including a first mark, and a relative position of the second area to the first area is detected using the template image. The method may further include a relative position detecting step of performing the following.

The image detecting method includes selecting a first area and a second area from a template image including a first mark and a second mark, respectively, and using the template image to determine a relative relation between the second area and the first area. The method may further include a relative position detection step of detecting a proper position.

The first area candidate portion detecting step includes a first projection signal forming step of forming a first projection signal by projecting a pixel value of the input image on a first axis, and a pixel value of the input image on a first axis. A second projection signal forming step of forming a second projection signal projected on a substantially vertical second axis; a first area first axis signal obtained by projecting a pixel value of the first area on a first axis; A first area first axis candidate position detecting step of detecting a position of the first area candidate portion in the first axis direction based on the first projection signal, and a first area having a pixel value of the first area projected on a second axis. The first region second detecting the position of the first region candidate portion in the second axis direction based on the region second axis signal and the second projection signal
Preferably, an axis candidate position detecting step is provided.

The second area candidate portion detecting step includes a third projection signal forming step of forming a third projection signal by projecting a pixel value of the second area candidate portion on a first axis, and a pixel value of the second area. Preferably, the method includes a third projection signal comparing step of comparing the second area first axis signal projected on the first axis with the third projection signal.

The second region candidate portion detecting step includes a fourth projection signal forming step of forming a fourth projection signal by projecting the pixel value of the second region candidate portion on a second axis, and a pixel value of the second region. The method may further include a fourth projection signal comparing step of comparing the second area second axis signal projected on the second axis with the fourth projection signal.

The first area first axis candidate position detecting step includes:
At different positions in the first axis direction, the first projection signal
Extracting a partial signal having a width of the region and comparing the partial signal with the first region first axis signal to calculate a first correlation value, based on the first correlation value, the first region in the first axis direction. The position of the candidate portion may be detected, and the first area second axis candidate position detecting step extracts the partial signal having the width of the first area from the second projection signal at a different position in the second axis direction, and The method may further include calculating a second correlation value by comparing with the one-region second-axis signal, and detecting a position of the first-region candidate portion in the second-axis direction based on the second correlation value.

In the third projection signal comparing step, a partial signal having a width of the second area is extracted from the third projection signal at a different position in the first axis direction, and each of the partial signals is compared with the first axis signal of the second area. The method may further include calculating a correlation value, wherein the image detection method includes detecting a first region approximation portion approximating the first region from the input image based on the third correlation value and the relative position. The method may further include a region approximate portion detecting step.

The first area first axis candidate position detecting step includes:
At different positions in the first axis direction, the first projection signal
Extracting a partial signal having a width of the region and comparing the partial signal with the first region first axis signal to calculate a first correlation value, based on the first correlation value, the first region in the first axis direction. The position of the candidate portion may be detected, and the first area second axis candidate position detecting step extracts the partial signal having the width of the first area from the second projection signal at a different position in the second axis direction, and Calculating a second correlation value by comparing with the one area second axis signal, and detecting a position of the first area candidate portion in the second axis direction based on the second correlation value; In the three-projection-signal comparing step, a partial signal having a width of the second area is extracted from the third projection signal at a different position in the first-axis direction, and the partial signal is compared with the first-axis signal of the second area to calculate a third correlation value The image detecting method may include the first step. At least one of the function value or the second correlation value, based a third correlation values, and the relative position, the first
The method may further include a first region approximation portion detecting step of detecting a first region approximation portion approximating the region from the input image.

In the third projection signal comparing step, a partial signal having a width of the second area is extracted from the third projection signal at a different position in the first axis direction, and is compared with the first axis signal of the second area. The fourth projection signal comparing step may include extracting a partial signal having a width of the second region from the fourth projection signal at a different position in the second axis direction, and calculating the correlation value. The method may include calculating a fourth correlation value by comparing the second correlation signal with the second axis signal. The image detection method may include calculating the first correlation value based on the third correlation value and the fourth correlation value.
The method may further include a first region approximation portion detecting step of detecting a first region approximation portion approximating the region from the input image.

The first area first axis candidate position detecting step includes:
A first area first axis signal edge extracting step of extracting an edge of the first area first axis signal; and a first projection signal edge extracting step of extracting an edge of the first projection signal.
The position of the first area candidate portion in the first axis direction may be detected based on the signal value of the edge of the area first axis signal and the signal value of the edge of the first projection signal, and the first area second axis The candidate position detecting step includes: a first area second axis signal edge extracting step of extracting an edge of the first area second axis signal;
A second projection signal edge extraction step of extracting an edge of the projection signal, wherein a second axis direction is determined based on a signal value of an edge of the first area second axis signal and a signal value of an edge of the second projection signal. May be detected.

The first area first axis candidate position detecting step includes:
The position of the first area candidate portion in the first axis direction may be detected based on the number of edges of the first area first axis signal and the number of edges of the first projection signal, and the first area second axis The candidate position detection step may detect the position of the first region candidate portion in the second axis direction based on the number of edges of the first region second axis signal and the number of edges of the second projection signal.

In the first area first axis candidate position detecting step,
Calculating a first edge correlation value indicating a correlation between the signal value of the edge of the first area first axis signal and the signal value of the edge of the first projection signal; and calculating the first edge correlation value based on the first edge correlation value.
The position of the first area candidate portion in the axial direction may be detected, and the first area second axis candidate position detecting step includes the step of detecting the signal value of the edge of the first area second axis signal and the signal of the edge of the second projection signal. Calculating a second edge correlation value indicating a correlation with the value, and detecting a position of the first region candidate portion in the second axis direction based on the second edge correlation value.

The third projection signal comparison step includes a second area first axis signal edge extraction step for extracting an edge of the second area signal, and a third projection signal edge extraction step for extracting an edge area of the third projection signal. Comparing the signal value of the edge of the second-axis first-axis signal with the signal value of the edge of the third projection signal.
The method may further include a first region approximate portion detecting step of detecting a first region approximate portion from the input image based on a signal value of an edge of the region first axis signal and a signal value of an edge of the third projection signal. The third projection signal comparison step may compare the number of edges of the second area first axis signal with the number of edges of the third projection signal.

The third projection signal comparing step includes a step of calculating a third edge correlation value indicating a correlation between an edge signal value of the second area first axis signal and an edge signal value of the third projection signal. The first area approximation portion detecting step may include:
The first area approximation part may be detected from the input image based on the third edge correlation value.

According to a second aspect of the present invention, there is provided an image detecting apparatus for detecting, from an input image, a first area approximation portion approximating a first area including a predetermined first mark. First region candidate portion detecting means for detecting a first region candidate portion including an approximated image from an input image, and information on an image of a second region having a known position relative to the first region are provided in advance. A second area candidate part detecting unit that determines a second area candidate part from the input image based on the relative position, and detects whether the second area candidate part matches information on the image of the second area. An image detecting device is provided.

According to a third aspect of the present invention, there is provided a wafer processing apparatus for exposing a circuit pattern on a wafer, wherein an image of a first area including a first mark and a relative position with respect to the first area are known. Storage means for storing information relating to the image of the second area; input image obtaining means for obtaining an image of the wafer as an input image; and a second area for detecting a first area candidate portion including an image similar to the first mark from the input image. A second region candidate portion is determined from the input image based on the one region candidate portion detection means and the relative position, and the second region candidate portion is determined by the second region candidate portion.
A first area approximation portion approximating the first region from the input image based on the detection result by the second region candidate portion detection means for detecting whether or not the information coincides with the information on the area image; Approximate part detection means for detecting
There is provided a wafer processing apparatus comprising: a position detecting unit that detects a position of a wafer based on a position of an approximate portion; and a moving unit that moves the wafer based on the detected position of the wafer.

Note that the above summary of the present invention does not list all of the necessary features of the present invention, and sub-combinations of these features may also constitute the present invention.

[0026]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described through embodiments of the present invention. However, the following embodiments do not limit the claimed invention and have the features described in the embodiments. Not all combinations are essential to the solution of the invention.

FIG. 1 is a block diagram showing a wafer processing apparatus according to one embodiment of the present invention. Wafer processing device 10
Has a template image storage unit 12, an input image acquisition unit 14, an image detection device 20, and a wafer moving unit 50. The wafer processing apparatus 10 may be, for example, an exposure apparatus that exposes a circuit pattern on a wafer or an inspection apparatus that inspects a pattern on a wafer. For example, in the case of an electron beam exposure apparatus, the wafer processing apparatus 10 further includes an electron gun that generates an electron beam, an electron lens that converges the electron beam and adjusts a focal position, and a deflection unit that deflects the electron beam. Is preferred.

The template image storage means 12 stores information on the template image including the first mark.
The template image storage unit 12 may further store information on a template image including a second mark whose position relative to the first mark is known. The template image stored in the template image storage unit 12 may be an image taken by an imaging device, or may be image data. Input image acquisition means 14
Acquires an image of a wafer as an input image.

The image detecting device 20 detects the position of the wafer based on the position of the first area approximation portion including the image approximated to the first mark in the input image. Image detection device 20
Means not only wafers, but also liquid crystal displays (LCD panels)
May be detected. The image detection device 20 includes a relative position detection unit 22, a first region candidate portion detection unit 24, a second region candidate portion detection unit 32,
It has a first area approximation portion detecting means 40 and a position detecting means 42.

The relative position detecting means 22 selects the first area including the first mark and the second area from the template image, and uses the information on the template image to determine the relative position of the second area to the first area. Detect the correct position.
It is preferable that the relative position detecting means 22 selects the first area and the second area such that the second area includes a second mark having a shape different from the first mark. Further, the relative position detecting means 22 may select an area that does not include the mark as the second area.

In another example, the wafer processing apparatus 10 does not include the relative position detecting means 22, and the template image storing means 12 may store information on the first area and the second area in advance. In this case, the first area includes the first mark. Further, the template image storage unit 12 stores a relative position of the second area with respect to the first area. Preferably, the second area includes a second mark having a shape different from the first mark. The second area may not have the mark.

The first region candidate portion detecting means 24 detects a first region candidate portion including an image approximate to the first mark from the input image. The first area candidate portion detecting means 24
It has an area signal forming unit 26, a projection signal forming unit 28, and a candidate position detecting unit 30. The first area signal forming means 26 converts the pixel value of the first area into the first area based on the template image stored in the template image storing means 12 and the position of the first area selected by the relative position detecting means 22. A first region first axis signal and a first region second axis signal projected on the first and second axes of the region, respectively, are formed.
Preferably, the second axis is substantially perpendicular to the first axis.
The projection signal forming means 28 includes a first projection signal and a second projection signal which project the pixel values of the input image on the first axis and the second axis, respectively.
Form the projection signal.

The candidate position detecting means 30 detects the position of the first region candidate portion in the first axis direction based on the first region first axis signal and the first projection signal. Further, the candidate position detecting means 30 detects the position of the first region candidate portion in the second axis direction based on the first region second axis signal and the second projection signal. The candidate position detecting means 30 detects the first mark candidate portion as described above.

The second region candidate portion detecting means 32 determines a second region candidate portion from the input image based on the relative position of the second region to the first region, and determines that the second region candidate portion is the second region. It is detected whether or not the information matches the information on the image. Specifically, the second area candidate portion detecting means 32 uses the image of the second mark included in the second area to detect whether or not the second area candidate portion matches information on the image of the second area. I do.

The second area candidate portion detecting means 32 has a second area signal forming means 34, a projection signal forming means 36, and a projection signal comparing means 38. Second area signal forming means 3
Reference numeral 4 forms a second-region first-axis signal and a second-region second-axis signal obtained by projecting the pixel values of the second region on the first axis and the second axis, respectively.

The projection signal forming means 36 is based on the position of the first area candidate portion detected by the candidate position detecting means 30 and the relative position of the second area detected by the relative position detecting means 22 with respect to the first area. Then, a second region candidate portion is determined from the input image. The projection signal forming means 36 forms a third projection signal and a fourth projection signal obtained by projecting the pixel values of the second area candidate portion on the first axis and the second axis, respectively.

The projection signal comparing means 38 compares the second area first axis signal with the third projection signal. Projection signal comparison means 3
8 further preferably compares the second area second axis signal with the fourth projection signal.

The first area approximation part detecting means 40 calculates the first area from the input image based on the detection result by the projection signal comparing means 38 and the position of the second area relative to the first area.
A first region approximation portion approximating the region is detected. In addition, the first area approximation part detecting means 40 includes the candidate position detecting means 30.
A first area approximation portion approximating the first area is detected from the input image based on both the detection result of the first area and the detection result of the projection signal comparing means 38 and the relative position of the second area to the first area. Is also good.

The position detecting means 42 detects the position of the wafer on the basis of the position of the approximate portion of the first area. The wafer moving means 50 moves the wafer based on the detected position of the wafer.

The wafer processing apparatus 10 according to the present embodiment
After detecting the first region candidate portion from the input image, a second region candidate portion is determined, and it is detected whether or not the second region candidate portion matches information on the image of the second region of the template image. Since the first region approximate portion is detected based on the first region, the first region approximate portion can be detected more accurately. Further, since the second area is an area obtained by dividing the template image, the second area is more similar to the case where the first area approximation part is detected based on whether or not the area corresponding to the template image of the input image matches information on the template image. , The first region candidate portion can be quickly detected.

FIG. 2 shows a wafer processing apparatus 10 according to the present invention.
5 is a flowchart showing each step of detecting a first area approximation portion from an input image.

In this embodiment, first, the relative position detecting means 22 selects the first and second regions from the template image, and detects the relative position of the second region with respect to the first region (S10). Next, the first area signal forming means 2
6 forms a first area first axis signal and a first area second axis signal (S12). Simultaneously with step 12 or step 1
Before or after Step 2, the projection signal forming means 28 forms the first projection signal and the second projection signal (S14). Subsequently, the candidate position detecting means 30 detects the position of the first region candidate portion from the input image (S16). Further, the second area signal forming means 34 forms a second area first axis signal and a second area second axis signal (S18). The first detected in step 16
Based on the position of the region candidate portion and the position of the second region detected in step 10 relative to the first region, the projection signal forming means 36 determines a second region candidate portion from the input image. Subsequently, the projection signal forming unit 36 forms a third projection signal and a fourth projection signal obtained by projecting the pixel values of the second area candidate portion (S20). The projection signal comparison means 38 detects whether or not the second area candidate portion matches the second area (S22). The first area approximation part detection means 40 detects the first area approximation part from the input image based on the detection result of the projection signal comparison means 38 (S24).

The details of each step will be described below with reference to the drawings. FIG. 3 is a schematic diagram illustrating a procedure for detecting a first region approximate portion from an input image of a wafer by the wafer processing apparatus 10 according to the embodiment of the present invention.

FIG. 3A shows a template image 100 given in advance. The template image 100 according to the present embodiment includes a first mark 102 and a second mark 104.
including. The first mark 102 has a line-and-space pattern in a predetermined direction.
04 has a line and space pattern in a direction different from that of the first mark 102. First mark 102 and second mark 102
The pattern of the mark 104 may be constituted by irregularities formed on the wafer by etching. In this case, for example, a CCD (Charge Coupled Device)
The line of the mark photographed by the imaging device such as e) has a different contrast from the space portion, and also has a different pixel value. The template image 100 may be image data formed based on the first mark 102 and the second mark 104.

In the present embodiment, since the first mark 102 and the second mark 104 have line and space patterns in different directions, the characteristics of the first mark 102 and the characteristics of the second mark 104 are respectively different. Projected on different axes. Therefore, even if the signal projected on any axis is affected by the reflected light on the wafer surface, the wafer processing apparatus 10 according to the present embodiment can accurately use the signal projected on the other axis. The first area approximation part can be detected.

As shown in FIG. 3B, the relative position detecting means 22 converts the first mark 10
2 and a second area 108 including the second mark 104 are selected. In another example, the relative position detection unit 22 may select an area 130 that does not include the mark as the second area. Further, the template image storage unit 12 may store information on the first area and the second area in advance. FIG. 3C shows the first region 106. The first area signal forming unit 26 is configured to project the pixel values of the first area on the first axis and the second axis, respectively.
An axis signal and a first area second axis signal are formed.

FIG. 3D shows the input image 110 obtained by the input image obtaining means 14. The projection signal forming unit 28 forms a first projection signal and a second projection signal that are obtained by projecting the input image on the first axis and the second axis, respectively. The pattern of the mark included in the input image 110 is formed by unevenness formed on the wafer by etching. Therefore, for example, an input image 1 captured by an imaging device such as a CCD
The line of the mark 10 has a different contrast from the space portion, and also has a different pixel value. Therefore, each of the first projection signal and the second projection signal has a pattern reflecting the mark pattern.

As shown in FIG. 3E, first, the candidate position detecting means 30 scans the first area first axis signal on the first projection signal. The candidate position detecting means 30 may extract a partial signal having a width of the first region from the first projection signal at a different position in the first axis direction, and compare each of the partial signals with the first region first axis signal. Specifically, the candidate position detecting means 30
A first area indicating a correlation between the first area first axis signal and the first projection signal;
The correlation value may be calculated, and the position of the first region candidate portion in the first axis direction may be detected based on the first correlation value.

In step 16 shown in FIG. 2, the candidate position detecting means 30 extracts a first area first axis signal edge for extracting an edge of the first area first axis signal.
Extracting a first projection signal edge for extracting an edge of the first projection signal. In this case, the candidate position detecting means 30 determines the position of the first region candidate portion in the first axis direction based on the signal value of the edge of the first region first axis signal and the signal value of the edge of the first projection signal. To detect. Specifically, the candidate position detecting means 30 calculates a first edge correlation value indicating a correlation between the signal value of the edge of the first area first axis signal and the signal value of the edge of the first projection signal, and The position of the first region candidate portion in the first axis direction may be detected based on the edge correlation value.

The first edge correlation value may be a normalized correlation value calculated based on the following equations (1) to (3). Equation 1 is an equation for calculating the signal value of the first area first axis signal.

(Equation 1) Here, T (m, n) is a pixel value of the first area at coordinates m in the first axis direction and coordinates n in the second axis direction. N is
This is the number of pixels in the second axis direction in the first region. T
x (m) is the first region first at the coordinate m in the first axis direction.
This is the signal value of the axis signal. The coordinates of the edge in the first area first axis signal are detected from the signal value Tx (m).

Equation 2 is an equation for calculating the signal value of the first projection signal.

(Equation 2) Here, X (i, j) is a pixel value at coordinates i in the first axis direction and coordinates j in the second axis direction of the input image. J is
This is the number of pixels in the second axis direction in the input image. X
x (i) is the signal value of the first projection signal at the coordinate i in the first axis direction. The coordinates of the edge in the first projection signal are detected from the signal value Xx (i).

Equation 3 is an equation for calculating a normalized correlation value.

(Equation 3) Here, refx is a reference data array indicating the coordinates of the edge in the first area first axis signal, and M ′ is the number of pixels of the edge detected from the first area first axis signal Tx (m). is there.

It is preferable that the candidate position detecting means 30 detect the coordinates on the first axis at which the first edge correlation value shows the maximum value as the first area first axis candidate part. In addition, the candidate position detection means 30 may detect, as coordinates in the first area and the first axis candidate portion, coordinates in which the first edge correlation value indicates the local maximum value and whose local maximum value is larger than a predetermined threshold value. In addition, the candidate position detecting means 30 determines, among the coordinates near the coordinates having the maximum value,
A coordinate having a first edge correlation value larger than a predetermined threshold value may be detected as the first area first axis portion.

The candidate position detecting means 30 calculates the coordinates on the first axis where the number of edges of the first area first axis signal and the number of edges of the first projection signal coincide with each other. May be detected. The candidate position detecting means 30 calculates the coordinates on the first axis where the first edge correlation value shows the maximum value and the number of edges of the first area first axis signal matches the number of edges of the first projection signal. It may be detected as one area first axis candidate part.

By the above operation, the candidate position detecting means 30
May detect a plurality of coordinates as the first area first axis candidate part. In the present embodiment, it is assumed that the value of the first correlation value between the partial signal 112 of the first projection signal and the first-axis first-axis signal has a maximum value. Therefore, the candidate position detecting means 30 detects the partial signal 112 as the first area first axis candidate part.

In this embodiment, the candidate position detecting means 3
In the case of 0, since the first region first axis candidate portion is detected using the edge portion of each signal, the first region candidate portion can be quickly detected.

In the present embodiment, the normalized correlation value is calculated using only the signal values of the coordinates of the edge of the first area first axis signal and the first projection signal. The normalized correlation value may be calculated according to the above equation.

Subsequently, the candidate position detecting means 30 scans the first area second axis signal on the second projection signal. The candidate position detecting means 30 detects the second position at a different position in the second axis direction.
A partial signal having the width of the first area may be extracted from the projection signal and compared with the first area second axis signal. Specifically, the candidate position detecting means 30 may calculate a second correlation value indicating a correlation between the first area second axis signal and the second projection signal, and calculate the second axis value based on the second correlation value. The position of the first area candidate portion in the direction may be detected.

In step 16 shown in FIG. 2, the candidate position detecting means 30 extracts a first area second axis signal edge for extracting an edge of the first area second axis signal.
Extracting a second projection signal edge for extracting an edge of the second projection signal. In this case, the candidate position detecting means 30 determines the position of the first area candidate portion in the second axis direction based on the signal value of the edge of the first area second axis signal and the signal value of the edge of the second projection signal. To detect. Specifically, the candidate position detecting means 30 calculates a second edge correlation value indicating a correlation between the signal value of the edge of the first area second axis signal and the signal value of the edge of the second projection signal, and The position of the first area candidate portion in the second axis direction may be detected based on the edge correlation value. As described above, the second edge correlation value may be a normalized correlation value calculated based on the same equation as the first edge correlation value.

It is preferable that the candidate position detecting means 30 detect the coordinates on the second axis at which the second edge correlation value shows the maximum value as the first area second axis candidate part. In addition, the candidate position detecting means 30 may detect, as coordinates in the first area and the second axis candidate part, coordinates in which the second correlation value indicates the local maximum value and the local maximum value is larger than a predetermined threshold value. In addition, candidate position detecting means 3
In the case of 0, among the coordinates near the coordinates having the maximum value, the coordinates having the second correlation value larger than the predetermined threshold value may be detected as the first area second axis portion.

The candidate position detecting means 30 calculates the coordinates on the second axis where the number of edges of the first area second axis signal and the number of edges of the second projection signal coincide with each other. May be detected. The candidate position detecting means 30 calculates the coordinates on the second axis where the second edge correlation value shows the maximum value and the number of edges of the first area second axis signal matches the number of edges of the second projection signal. It may be detected as one area second axis candidate part.

By the above operation, the candidate position detecting means 30
May detect a plurality of coordinates as the first area second axis candidate part. In the present embodiment, it is assumed that the value of the second correlation value between the partial signal 114 of the second projection signal and the first area second axis signal indicates a local maximum value. Therefore, the candidate position detecting means 30 detects the partial signal 114 as the first area second axis candidate part.

In this embodiment, the candidate position detecting means 3
In the case of 0, since the first area second axis candidate part is detected using the edge part of each signal, the first area candidate part can be quickly detected.

The candidate position detecting means 30 detects a region specified by the partial signals 112 and 114 detected as described above as a first region candidate portion 122. The projection signal forming unit 36 determines the first region candidate portion 122 detected by the candidate position detecting unit 30 and the relative position of the second region detected by the relative position detecting unit 22 with respect to the first region. A second region candidate portion 124 is determined.

FIG. 3F shows the second region 108. The second area signal forming means 34 forms a second area first axis signal and a second area second axis signal by projecting the pixel values of the second area on the first axis and the second axis, respectively.

FIG. 3G shows the second area candidate portion 124 determined from the input image 110. The projection signal forming means 36 forms a third projection signal and a fourth projection signal obtained by projecting the pixel values of the second region candidate portion 124 on the first axis and the second axis, respectively.

First, the projection signal comparing means 38 scans the second area first axis signal on the third projection signal. The projection signal comparing means 38 may extract a partial signal having a width of the second region from the third projection signal at a different position in the first axis direction, and compare each of the partial signals with the second region first axis signal. Specifically, the projection signal comparing means 38 may calculate a third correlation value indicating a correlation between the second area first axis signal and the third projection signal, and calculate the second area based on the third correlation value. It may be detected whether or not the candidate portion matches information on the image of the second area.

In step 22 shown in FIG. 2, the projection signal comparing means 38 extracts a second area first axis signal edge extracting step of extracting an edge of the second area first axis signal, and executes an edge area of the third projection signal. And extracting a third projection signal edge to be extracted. In this case, the projection signal comparing means 38 compares the signal value of the edge of the second region first axis signal with the third signal value.
The signal value of the edge of the projection signal is compared. In particular,
The projection signal comparison means 38 may calculate a third edge correlation value indicating a correlation between the signal value of the edge of the second area first axis signal and the signal value of the edge of the third projection signal.

As described above, the third edge correlation value may be a normalized correlation number calculated based on the same equation as the first edge correlation value. It is preferable that the projection signal comparing means 38 detect a coordinate having a third edge correlation value larger than a predetermined threshold value as a second area first axis portion. Further, the projection signal comparing means 38 may detect the coordinates on the first axis at which the third edge correlation value has the maximum value as the second area first axis candidate part. Further, the projection signal comparing means 38 may detect, as coordinates in the second area first axis candidate, coordinates in which the third edge correlation value indicates the maximum value, in which the maximum value is larger than a predetermined threshold value. Further, the projection signal comparing means 38 may detect, as coordinates in the second area first axis candidate portion, coordinates having a third correlation value larger than a predetermined threshold value among coordinates near the coordinates having the maximum value.

Further, the projection signal comparing means 38 calculates the coordinates on the first axis where the number of edges of the second area first axis signal coincides with the number of edges of the third projection signal, in the second area first axis candidate portion. May be detected. The projection signal comparing means 38 calculates the coordinates on the first axis where the number of edges of the second area first axis signal and the number of edges of the third projection signal coincide with each other, and the third edge correlation value shows the maximum value. It may be detected as a two-region first axis candidate portion.

With the above operation, the projection signal comparing means 38
May detect a plurality of coordinates as the second area first axis candidate part. In the present embodiment, the projection signal comparing means 38
Uses the edge portion of each signal to detect the second region first axis candidate portion, so that the second region candidate portion
It can be detected whether or not the information matches the information on the image of the area.

The first area approximation part detecting means 40 determines the position of the second area candidate part specified by the second area first axis candidate part thus detected and the relative position of the second area to the first area. A first region approximation portion approximating the first region is detected from the input image based on the general position.

The projection signal comparing means 38 may further scan the second area second axis signal on the fourth projection signal and compare the second area second axis signal with the fourth projection signal. The projection signal comparing means 38 extracts the partial signal having the width of the second area from the fourth projection signal at a different position in the second axis direction,
Each may be compared with the second area second axis signal. Specifically, the projection signal comparing unit 38 may calculate a fourth correlation value indicating a correlation between the second area second axis signal and the fourth projection signal, and the projection signal comparing unit 38 may calculate the fourth correlation value. Whether or not the second area candidate portion matches information on the image of the second area may be detected based on the value.

In step 22 shown in FIG. 2, the projection signal comparing means 38 extracts the edge of the second area second axis signal edge to extract the edge of the second area second axis signal, and extracts the edge area of the fourth projection signal. And extracting a fourth projection signal edge to be extracted. In this case, the projection signal comparing means 38 compares the signal value of the edge of the second area second axis signal with the fourth signal.
The signal value of the edge of the projection signal is compared. In particular,
The projection signal comparison means 38 may calculate a fourth edge correlation value indicating a correlation between the signal value of the edge of the second area second axis signal and the signal value of the edge of the fourth projection signal. As described above, the fourth edge correlation value may be a normalized correlation number calculated based on the same equation as the first edge correlation value.

It is preferable that the projection signal comparing means 38 detect a coordinate having a fourth edge correlation value larger than a predetermined threshold value as the second area second axis portion. Further, the projection signal comparing means 38 may detect the coordinates on the second axis at which the fourth edge correlation value indicates the maximum value as the second area second axis candidate part.
In addition, the projection signal comparing means 38 may detect, as coordinates in the second area second axis candidate portion, coordinates in which the fourth edge correlation value indicates the maximum value, in which the maximum value is larger than a predetermined threshold value.
Further, the projection signal comparing means 38 may detect, as coordinates in the second area second axis candidate portion, coordinates having a fourth correlation value larger than a predetermined threshold value among coordinates near coordinates having the maximum value.

The projection signal comparing means 38 calculates the coordinates on the second axis where the number of edges of the second area second axis signal and the number of edges of the fourth projection signal coincide with each other, in the second area second axis candidate portion. May be detected. The projection signal comparing means 38 calculates the coordinates on the second axis where the fourth edge correlation value indicates the maximum value and the number of edges of the second area second axis signal matches the number of edges of the second projection signal. It may be detected as a two-region second axis candidate portion. By the above operation, the candidate position detecting means 30 may detect a plurality of coordinates as the second area second axis candidate part.

The projection signal comparing means 38 detects whether or not the second area candidate portion matches the information on the image of the second area based on both the third edge correlation value and the fourth edge correlation value. Good.

The first area approximation portion detecting means 40 calculates at least one of the first edge correlation value and the second edge correlation value,
A first area approximation portion approximating the first area may be detected from the input image based on the third edge correlation value and the relative position of the second area to the first area. For example, the first area approximation part detection means 40 may detect the first area approximation part from the input image based on the third edge correlation value and the relative position of the second area to the first area. Also, the first
The region approximate portion detection means 40 may detect the first region approximate portion from the input image based on the third edge correlation value and the fourth edge correlation value and the relative position of the second region to the first region. Good.

In another example, the relative position detecting means 22
May select the entire template image 100 as the first area and select only the area including the second mark included in the template image 100 as the second area.

In another example, the first area candidate portion may be detected using an amount for evaluating a difference between two images, such as a difference value between the template image 100 and the input image 110. Further, in order to calculate the normalized correlation value, the pixel value of the template image 100 and the pixel value of the input image 110 may be two-dimensionally compared.

FIG. 4 shows the first region 106 shown in FIG.
FIG. 7 is a schematic diagram showing a first area second axis signal obtained by projecting a second axis on a second axis. In FIG. 4A, the first axis is the horizontal axis, and the second axis is the second axis.
The axis is the vertical axis. In another example, the first axis is the Y axis,
The second axis may be the X axis, and the first and second axes may be in any direction.

FIG. 4 (b) shows the first area when the signal value of the first area second axis signal projected on the second axis is plotted on the horizontal axis and the coordinate in the second axis direction is plotted on the vertical axis. 2 shows a two-axis signal. The first area second axis signal indicates a pattern reflecting the pattern of the first mark included in the first area. In the present embodiment, the first mark has a pattern including four lines parallel to the first axis direction. Therefore, the first area second axis signal has edges reflecting four lines parallel to the first axis.

FIG. 5 is a chart showing steps for extracting an edge from each signal. The first area signal forming means 26 differentiates the first area first axis signal to calculate a primary differential value. It is preferable that the first area signal forming means 26 extracts, as an edge, a coordinate at which the absolute value of the differentiated primary differential value is larger than a predetermined value. Further, the first area signal forming means 26 further differentiates the primary extreme value coordinates where the differentiated primary differential value becomes an extreme value, and the differentiated secondary differential value sandwiching the primary extreme value coordinates becomes a minimum. The edge from the coordinates to the coordinates at which the second derivative value becomes maximum may be extracted as an edge.

FIG. 5A to FIG. 5C show steps of detecting an edge region of a falling portion in the present embodiment. FIG. 5A shows signal values for the coordinates of the first area first axis signal. FIG. 5B shows the first derivative and the second derivative with respect to the coordinates of the first area first axis signal. The first area signal forming means 26 detects primary extreme value coordinates at which the primary differential value becomes a minimum value. As shown in FIG. 5 (c), the first area signal forming means 26 calculates a range from the coordinate at which the differentiated secondary differential value sandwiching the primary extreme value coordinate becomes minimum to the coordinate at which the secondary differential value becomes maximum. It is extracted as the falling edge of the 1-axis first axis signal.

FIGS. 5D to 5F show steps of detecting the edge of the rising portion in this embodiment. FIG. 5D shows a signal value with respect to the coordinates of the first area first axis signal. FIG. 5E shows the first derivative and the second derivative with respect to the coordinates of the first area first axis signal. The first area signal forming means 26 detects primary extreme value coordinates at which the primary differential value becomes a local maximum value. As shown in FIG. 5 (f), the first area signal forming means 26 calculates the differentiated 2 with respect to the primary extreme value coordinate.
The coordinates from the coordinates at which the second derivative value is maximum to the coordinates at which the second derivative value is minimum are extracted as rising edges of the first area first axis signal.

The edges of the first projection signal, the second projection signal, the third projection signal, the fourth projection signal, the second area first axis signal, and the second area second axis signal are detected in the same manner. Is preferred.

FIG. 6 is a schematic diagram showing an example of a template image stored in the template image storage means 12 of the wafer processing apparatus 10 according to the present embodiment. The template image 200 includes three marks 202, 204 and 206, as shown in FIG. Relative position detecting means 2
2 is a template image 20 as shown in FIG.
0 is divided into four regions 210, 212, 214 and 216. The relative position detection unit 22 sets the area 210 including the mark 210 as a first area and the area 214 including the mark 206 as a second area.

FIG. 6C shows an input image 220 of the wafer acquired by the input image acquiring means 14. Input image 220
In, the region 222 has a mark substantially equal to the template image 200. For example, the mark corresponding to the mark 204 in the area 222 is a dark mark 22 as a whole.
4, the pixel value of the area 222 is set to the first
The signal values of the projection signals projected in the axial direction and the second axial direction are:
The pixel values of the template image 200 are set in the first axis direction and the second
It differs from the signal value of the template signal projected in the axial direction.
Therefore, the position of the region 222 cannot be accurately detected.

On the other hand, in the present embodiment, FIG.
As shown in FIG. 5, the region 222 is a region 22 corresponding to the first region and the second region selected from the template image 200.
6 and 228, and the template image 2
Compare the image of the first area and the image of the second area of 00. In this case, the signal values of the projection signals obtained by projecting the pixel values of the regions 226 and 228 in the first axis direction and the second axis direction, respectively, do not include the pixel value of the deformed mark 224. Therefore, the wafer processing apparatus 10 according to the present embodiment can accurately detect the position of the region 222.

FIG. 7 is a schematic diagram showing an example of a combination of the first area and the second area of the template image selected by the relative position detecting means 22 in the wafer processing apparatus 10 according to the present embodiment.

As shown in FIG. 7A, the relative position detecting means 22 may select the entire template image 100 as the first area and select the shaded area as the second area.

Further, the relative position detecting means 22 is provided as shown in FIG.
As shown in (b) and FIG. 7 (c), the template signal 100 may be divided into four, and the area indicated by oblique lines may be set as the first area, and other areas may be set as the second areas.

Further, the relative position detecting means 22 of the area obtained by dividing the template image 100 into four
Only two regions shown by oblique lines in (b) may be selected as the first region and the second region. In addition, the relative position detection unit 22 may select only the three areas indicated by oblique lines in FIG. 7C as the first area and the second area from the area obtained by dividing the template image 100 into four. . in this case,
The relative position detecting means 22 may select the third area as the third area. The relative position detecting means 22 detects the relative positions of the second area and the third area with respect to the first area. The first area candidate portion detecting means 24 detects, from the input image, a first area candidate portion including an image similar to the first mark included in the first area. Second area candidate portion detecting means 3
2 determines a second region candidate portion and a third region candidate portion from the input image based on the relative positions of the second region and the third region with respect to the first region, respectively. It is detected whether or not each of the candidate portions matches the information on the image in the second area and the information on the image in the third area. The first area approximation part detection means 40 detects the first area approximation part from the input image based on the detection result of the second area candidate part detection means 32 regarding the second area candidate part and the third area candidate part.

Similarly, the relative position detecting means 22 is provided as shown in FIG.
As shown in (d), the template image 100 is divided into four parts, and the first area, the second area,
The third region and the fourth region may be selected. in this way,
By performing detection using a plurality of regions, the first region approximation portion can be detected more accurately.

The relative position detecting means 22 divides the template image 100 into nine, as shown in FIG.
The hatched area may be selected as the first area, and the other area may be selected as the second area.

As described above, the relative position detecting means 2
2 may use any combination of marks in the template image as the first area and the second area. Also, not all marks in the template image need be used as the first area or the second area.

FIG. 8 is a schematic diagram showing an example of the input image obtained by the input image obtaining means 14. Input image 3
00 is the alignment mark 302 and the circuit pattern 30
4 inclusive. The template image storage unit 12 may store the alignment mark 302 on the wafer as a first mark and the shape of the alignment circuit pattern 304 as a second mark. In this case, the template image storage unit 12 also stores the position of the area including the circuit pattern relative to the alignment mark.

The wafer processing apparatus 10 according to the present embodiment
After detecting the first region candidate portion from the input image, a second region candidate portion is determined, and it is detected whether or not the second region candidate portion matches information on the image of the second region of the template image. Since the first region approximate portion is detected based on the first region, the first region approximate portion can be detected more accurately.

Since the second region is a region obtained by dividing the template image, the first region approximation portion is detected based on whether or not the region of the input image corresponding to the template image matches the information on the template image. The first region candidate portion can be detected more quickly than in the case.

Further, in the wafer processing apparatus 10 according to the present embodiment, since the marks in the template image can be variously combined to form the first area and the second area, the influence of the reflected light on the wafer surface does not occur. By selecting a combination of marks, the first area approximation portion can be accurately detected.

As described above, the present invention has been described using the embodiment. However, the technical scope of the present invention is not limited to the scope described in the above embodiment. Various changes or improvements can be added to the above embodiment. It should be noted that such modified or improved embodiments may be included in the technical scope of the present invention.
It is clear from the description of the claims.

[0102]

As is clear from the above description, according to the present invention, a mark substantially the same as the template image is formed.
It can be accurately detected from the input image.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a wafer processing apparatus according to one embodiment of the present invention.

FIG. 2 is a flowchart showing each step of detecting a first area approximation portion from an input image by the wafer processing apparatus according to the embodiment.

FIG. 3 is a schematic diagram showing a procedure for detecting a first area approximation portion from an input image of a wafer by a wafer processing apparatus according to an embodiment of the present invention.

FIG. 4 is a schematic diagram showing a first area first axis signal and a first area second axis signal obtained by projecting an image of a first area onto a first axis and a second axis.

FIG. 5 is a chart showing a procedure for extracting an edge from each signal.

FIG. 6 is a schematic diagram illustrating an example of a template image stored in a template image storage unit of the wafer processing apparatus according to the present embodiment.

FIG. 7 is a schematic diagram illustrating an example of a combination of a first area and a second area of a template image selected by a relative position detection unit in the wafer processing apparatus according to the present embodiment.

FIG. 8 is a schematic diagram illustrating an example of an input image acquired by an input image acquiring unit of the wafer processing apparatus according to the present embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 10 ... Wafer processing apparatus, 12 ... Template image storage means, 14 ... Input image acquisition means, 20 ... Image detection apparatus, 2
2 ... relative position detecting means, 24 ... first area candidate part detecting means, 26 ... first area signal forming means, 28 ... projection signal forming means, 30 ... candidate position detecting means, 32 ... second area candidate part detecting means, 34: second area signal forming means, 36: projection signal forming means, 38: projection signal comparing means, 40: first area approximating portion detecting means, 42: position detecting means, 50: wafer moving means, 100: template image, 102 ... first
Mark, 104: second mark, 106: first area, 10
8: second region, 110: input image, 122: first region candidate portion, 124: second region candidate portion

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06T 7/60 150 G06T 7/60 150B H01L 21/027 H01L 21/30 541K F-term (Reference) 5B057 AA03 BA02 CA12 CA16 DA07 DB02 DC16 DC19 5F056 BC10 BD01 BD05 5L096 BA03 CA02 FA06 FA38 FA69 HA08 JA03 JA09

Claims (20)

[Claims] 1. An image detection method for detecting, from an input image, a first region approximation portion approximating a first region including a predetermined first mark, wherein the first region includes an image approximation to the first mark. A first area candidate part detecting step of detecting an area candidate part from the input image; and information about an image of a second area whose relative position with respect to the first area is known is given in advance, and Determining a second area candidate part from the input image based on the position, and detecting whether the second area candidate part matches information on the image of the second area. An image detection method, comprising: 2. The method according to claim 1, wherein the second area candidate portion detecting step includes:
2. The method according to claim 1, wherein it is detected whether or not the second area candidate portion matches information on the image of the second area using an image of a second mark included in the second area. 3. Image detection method. 3. The step of detecting a second area candidate portion,
Detecting whether or not the second area candidate portion matches information on the image of the second area using an image of a second mark included in the second area and having a shape different from the first mark. The image detection method according to claim 1, wherein: 4. The method according to claim 1, wherein the first area and the second area are respectively selected from a template image including the first mark, and the relative position of the second area to the first area is selected using the template image. 2. The method according to claim 1, further comprising the step of detecting a relative position.
2. The image detection method according to 1., 5. The method according to claim 1, wherein the first area and the second area are respectively selected from a template image including the first mark and the second mark, and the first area of the second area is selected using the template image. The image detection method according to claim 2, further comprising a relative position detection step of detecting the relative position with respect to an area. 6. The first area candidate portion detecting step includes: a first projection signal forming step of forming a first projection signal obtained by projecting a pixel value of the input image on a first axis; and a pixel value of the input image. Forming a second projection signal by projecting on a second axis substantially perpendicular to the first axis; and a second step of projecting the pixel value of the first area on the first axis. A first area first axis candidate position detecting step of detecting a position of the first area candidate portion in the first axis direction based on a first area first axis signal and the first projection signal; A first region detecting a position of the first region candidate portion in the second axis direction based on a first region second axis signal and a second projection signal obtained by projecting a pixel value on the second axis. 2. A two-axis candidate position detecting step. Image detecting method according to any one of al 5. 7. The second area candidate part detecting step includes: forming a third projection signal that projects a pixel value of the second area candidate part on the first axis; 7. A third projection signal comparing step of comparing a second area first axis signal obtained by projecting pixel values of two areas on the first axis with the third projection signal.
2. The image detection method according to 1., 8. The second area candidate part detecting step includes: forming a fourth projection signal that projects a pixel value of the second area candidate part on the second axis; 8. The method according to claim 7, further comprising: a fourth projection signal comparing step of comparing a second area second axis signal obtained by projecting pixel values of two areas on the second axis with the fourth projection signal. The image detection method described in the above. 9. The first area first axis candidate position detecting step extracts a partial signal having a width of the first area from the first projection signal at a different position in the first axis direction,
Calculating a first correlation value by comparing the first correlation value with the first region first axis signal, and detecting a position of the first region candidate portion in the first axis direction based on the first correlation value The first area second axis candidate position detecting step includes:
Extracting a partial signal having a width of the first area from the second projection signal at a different position in the axial direction, and comparing the partial signal with the first area second axis signal to calculate a second correlation value, 9. The image detection method according to claim 6, wherein a position of the first area candidate portion in the second axis direction is detected based on the second correlation value. 10. The third projection signal comparing step includes extracting partial signals having a width of the second area from the third projection signal at different positions in the first axis direction, and respectively extracting the second area first axis. Calculating a third correlation value by comparing with a signal, detecting a first area approximation portion approximating the first area from the input image based on the third correlation value and the relative position. The image detection method according to claim 7, further comprising a first area approximation portion detection step of performing the following. 11. The first area first axis candidate position detecting step includes detecting the first area at a different position in the first axis direction.
Extracting a partial signal having a width of the first area from the projection signal, and calculating a first correlation value by comparing the partial signal with the first area first axis signal; based on the first correlation value, Detecting a position of the first area candidate portion in the first axis direction; and detecting the first area second axis candidate position in the first area.
Extracting a partial signal having a width of the first area from the second projection signal at a different position in the axial direction, and comparing the partial signal with the first area second axis signal to calculate a second correlation value, Detecting a position of the first area candidate portion in the second axis direction based on the second correlation value; and comparing the third projection signal at a different position in the first axis direction. Extracting the partial signal having the width of the second region from the first region, and comparing the partial signal with the first axis signal of the second region to calculate a third correlation value, wherein the first correlation value or the second correlation value is calculated. At least one of
A first area approximation portion detecting step of detecting a first area approximation portion approximating the first region from the input image based on the third correlation value and the relative position. The image detection method according to claim 7. 12. The third projection signal comparing step includes extracting partial signals having a width of the second area from the third projection signal at different positions in the first axis direction, and respectively extracting the second area first axis. Calculating a third correlation value by comparing with a signal, wherein the fourth projection signal comparing step includes:
Extracting a partial signal having a width of the second region from the fourth projection signal at a different position in the second axis direction and comparing the partial signal with the second region second axis signal to calculate a fourth correlation value. A first area approximation part detecting step of detecting a first area approximation part approximating the first area from the input image based on the third correlation value and the fourth correlation value. The image detection method according to claim 7, further comprising: 13. The first area first axis candidate position detecting step includes: a first area first axis signal edge extracting step of extracting an edge of the first area first axis signal; and an edge of the first projection signal. Extracting a first projection signal edge in the first axis direction based on a signal value of an edge of the first area first axis signal and a signal value of an edge of the first projection signal. Detecting the position of the first area candidate portion, and detecting the first area second axis candidate position, wherein the first area second axis signal extracting section extracts an edge of the first area second axis signal.
An axis signal edge extraction step; and a second projection signal edge extraction step of extracting an edge of the second projection signal, wherein a signal value of an edge of the first area second axis signal and an edge of the second projection signal 7. The image detection method according to claim 6, wherein the position of the first area candidate portion in the second axis direction is detected based on the signal value of the first region. 14. The first area first axis candidate position detecting step includes the step of detecting the first area first axis direction based on the number of edges of the first area first axis signal and the number of edges of the first projection signal. Detecting the position of the first area candidate portion in the first area, the first area second axis candidate position detecting step includes determining the number of edges of the first area second axis signal and the number of edges of the second projection signal. 14. The image detection method according to claim 13, wherein a position of the first region candidate portion in the second axis direction is detected based on the position. 15. The first area first axis candidate position detecting step includes a first edge indicating a correlation between a signal value of an edge of the first area first axis signal and a signal value of an edge of the first projection signal. Calculating a correlation value, detecting a position of the first area candidate portion in the first axis direction based on the first edge correlation value, and detecting the first area second axis candidate position, Calculating a second edge correlation value indicating a correlation between a signal value of an edge of the first area second axis signal and a signal value of an edge of the second projection signal, based on the second edge correlation value The method according to claim 13, wherein the position of the first area candidate portion in the second axis direction is detected. 16. The third projection signal comparison step includes a second area first axis signal edge extraction step for extracting an edge of the second area signal, and a third projection for extracting an edge area of the third projection signal. A signal edge extraction step, and a step of comparing a signal value of an edge of the second area first axis signal with a signal value of an edge of the third projection signal, wherein the edge of the second area first axis signal is 8. A first area approximation portion detecting step of detecting the first area approximation portion from the input image based on the signal value of the third projection signal and the signal value of the edge of the third projection signal. 3. The image detection method according to 1., 17. The method according to claim 16, wherein the third projection signal comparing step compares the number of edges of the second area first axis signal with the number of edges of the third projection signal.
2. The image detection method according to 1., 18. The third projection signal comparing step includes calculating a third edge correlation value indicating a correlation between an edge signal value of the second area first axis signal and an edge signal value of the third projection signal. Performing the first area approximation portion detection step, wherein the first area approximation portion detection step includes:
18. The image detection method according to claim 16, wherein a region approximate portion is detected from the input image. 19. The method according to claim 1, further comprising the step of:
An image detecting apparatus for detecting, from an input image, a first area approximation portion approximating an area, wherein a first area candidate portion including an image approximating the first mark is detected from the input image. Means, and information on an image of a second area whose relative position to the first area is known is given in advance, and a second area candidate portion is determined from the input image based on the relative position. An image detecting apparatus, comprising: a second area candidate detecting unit configured to detect whether the second area candidate matches information on an image of the second area. 20. A wafer processing apparatus for exposing a circuit pattern on a wafer, comprising: an image of a first area including a first mark and information of an image of a second area having a known position relative to the first area. Storage means for storing an image of the wafer, an input image obtaining means for obtaining an image of the wafer as an input image, and a first area candidate part for detecting a first area candidate part including an image similar to the first mark from the input image Detecting means for determining a second region candidate portion from the input image based on the relative position, and detecting whether or not the second region candidate portion matches information on the image of the second region. 2
An approximate part detecting means for detecting a first area approximate part approximating the first area from the input image based on a detection result by the area candidate part detecting means and the second area candidate part detecting means; A wafer processing apparatus comprising: a position detecting unit that detects a position of the wafer based on a position; and a moving unit that moves the wafer based on the detected position of the wafer.