JPS60177647A - Automatic precision positioning system - Google Patents

Abstract

PURPOSE:To enable to precisely position a wafer by a method wherein the relative position of the image of the wafer and the linear region in a memory are detected by a system other than a pattern matching system and the wafer is subjected to a primary positioning, and after that, the relative position is detected by the pattern matching system and the wafer is subjected to a secondary positioning. CONSTITUTION:A wafer 2 is held on a movable stand 4, part of the surface of the wafer 2 is enlarged 26 and picked up, the concentration signal of lattice state array picture elements in the image is undergone an A/D conversion 32 and the concentration signal converted is stored in an image frame memory 34. An arithmetic processing 40 of a differentiation 44, a threshold processing and a parity check 48 is performed to the memory 34 to make a binary digital signal, which is memorized in a CPU36, and the relative position of a linear region 18 to the image is detected 42 by said signal. Successively, when the wafer is remaining unchanged on the prescribed position, a key pattern 56, a subkey pattern 57 and the positions thereof, which are displayed on each left and right part of a display screen 50, are memorized 52. The pattern in the image same as the key pattern is detected 54 on the basis of the memories 34 and 52, the matching degree is calculated 36 and the linear region 18 is made to match while the driving 21 of the stand 4 is controlled 64. According to this constitution, the automatic positioning of the wafer can be performed fully, rapidly and accurately.

Description

[Detailed Description of the Invention] [Technical Field] The present invention relates to an automatic precision alignment system for semiconductor wafers, and more specifically, the present invention relates to an automatic precision alignment system for semiconductor wafers. The present invention relates to an automatic precision positioning system for positioning a semiconductor wafer, each of which has a circuit pattern in a plurality of rectangular regions defined by linear regions, at a desired position.

<Background Art> As is well known, on the surface of a semiconductor wafer, there are a plurality of linear regions having a predetermined width and arranged in a lattice shape at predetermined intervals. Such linear regions are generally referred to as streets. A circuit pattern is applied to a plurality of rectangular areas partitioned by linear areas. Such a semiconductor wafer is cut in the linear region, and a plurality of rectangular regions each having a circuit pattern are thus separated. Individually separated rectangular areas are commonly referred to as chips. It is important to cut the semiconductor wafer with sufficient precision in the linear region, and the width of the linear region itself is extremely narrow, generally on the order of several tens of micrometers. Therefore, when cutting a semiconductor wafer with a TK cutting means, such as a diamond blade, it is necessary to align the semiconductor wafer with great precision with respect to the cutting means.

Various forms of automatic precision positioning systems have already been proposed and put into practical use for positioning the semiconductor wafer at a desired position with sufficient precision for the above-mentioned cutting and the like. Such automatic precision alignment systems generally detect with sufficient precision the relative position of the linear region on the surface of the semiconductor wafer held by the holding means, and move the holding means based on such detection. Semiconductor quaeno 1
is aligned to the required position. Detection of the relative position of the linear region in such an automatic precision alignment system generally utilizes a pattern matching method. That is, a pattern of characteristic specific areas on the surface of the semiconductor wafer when it is in a predetermined position, that is, a key pattern;
The position of the key pattern is stored in advance, and the same pattern as the key pattern is detected on the surface of the semiconductor wafer to be aligned, thus detecting the relative position of the linear region.

However, conventional automatic precision alignment systems that utilize pattern matching methods to detect the relative positions of linear regions do not require comparison on the surface of a semiconductor wafer to properly detect the relative positions of linear regions. There is a serious drawback in that pattern matching must be performed over a wide range of targets, thus requiring a considerable amount of time, thus hindering the speeding up of semiconductor wafer processing steps such as cutting. .

On the other hand, Patent Application No. 16 filed in 1982 by the present applicant
2031, title of invention: automatic precision positioning system, filing date: September 5, 1988) The specification and drawings state that the position of at least one edge of the linear region is detected by a unique straight line detection method; Thus, automatic precision alignment systems have been proposed that utilize detection of the relative positions of the linear regions. This automatic precision alignment system is capable of detecting the relative position of a linear region considerably more quickly than the conventional automatic precision alignment system that uses a pattern matching method, and therefore, it is possible to detect the relative position of a linear region considerably more quickly than the conventional automatic precision alignment system that uses a pattern matching method. It is possible to sufficiently cope with speeding up the processing process.

However, although the automatic precision alignment system proposed in the specification and drawings of Patent Application No. 162031 of 1988 has a very small possibility, If there are many straight lines that are similar to the side edges of a linear region in the circuit pattern being applied, such straight lines may be erroneously detected as the side edges of the linear region, thus causing errors in the positioning of the semiconductor wafer. It turned out that there was a fear.

<Object of the Invention> The present invention has been made in view of the above facts, and its main purpose is to substantially eliminate the occurrence of errors and to position a semiconductor wafer at a desired position sufficiently quickly. , to provide an improved automatic precision alignment system.

<Minimum of the Invention> As a result of intensive study, the present inventors have determined that a non-pattern matching method, preferably the above-mentioned Patent Application No. 162031 filed in 1982,
Detection of the relative position of a linear region by the continuous independent linear detection method proposed in the specification and drawings of the patent,
When combined with the detection of the relative position of a linear region by the putter/matching method, it is possible to compensate for the occurrence of false spots in the former detection in the latter detection, and the former detection can also be used for the latter detection. It has been found that the time taken to perform the process can be considerably shortened, thus making it possible to substantially eliminate the occurrence of errors and positioning the semiconductor wafer at the desired position sufficiently quickly.

That is, according to the present invention, there are a plurality of linear regions arranged in a grid on the surface, and a circuit pattern is applied to each of the plurality of rectangular regions partitioned by the linear regions. a holding means for holding the semiconductor wafer; a moving means for moving the holding means; and a holding means for moving the holding means; imaging means for capturing an image of at least a portion of the surface of the semiconductor wafer 71 and outputting an analog signal indicating the density of the x-y matrix array image; an image frame memory for storing signals, a key pattern memory for storing a key pattern corresponding to a specific area on the surface when the semiconductor wafer is in a predetermined position, and a signal indicating the position of the key pattern; Detecting a pattern identical to the key pattern in the image captured by the imaging means based on the signal stored in the image frame memory and the key pattern signal stored in the key pattern memory. a pattern matching type position detection means for detecting the relative position of the linear area, and a non-pattern matching type position detection means for detecting the relative position of the linear area by a method other than pattern matching. position detecting means; and actuating the moving means in response to detection of the relative position of the linear region by the non-pattern matching position detecting means, thus performing primary positioning of the semiconductor wafer held by the holding means. Thereafter, the moving means is actuated in accordance with the detection of the relative position of the linear region by the pattern matching type position detection means, thereby secondary positioning of the semiconductor wafer held by the holding means. An automatic precision alignment system is provided, comprising a movement control means for performing the following steps.

<Preferred Specific Example of the Invention> Hereinafter, a specific example of an automatic precision alignment system configured according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 schematically depicts a portion of a semiconductor wafer cutting apparatus equipped with an embodiment of an automatic precision alignment system constructed in accordance with the present invention.

The semiconductor wafer/2 to be cut is placed on the holding means 4, fed by suitable feeding means (not shown), which may be of a form known per se.

At this time, for example, by using the orientation flat 6 existing on the wafer 2, the wafer 2 is placed on the holding means 4 within a required error range, although not with sufficient precision.

To explain this point in more detail, as shown in FIG. 2, there are a plurality of linear regions 8 arranged in a grid pattern on the surface of the wafer 20. These linear regions 8, generally called streets, have a predetermined width W and are arranged at predetermined intervals d. The width of the linear region 8a extending in a specific direction and the width of the linear region 8b extending in a direction perpendicular to the specific direction do not necessarily have to be substantially the same and are often different from each other, Generally, it is about several tens of μm. Further, the spacing between the linear regions 8a extending in the specific direction and the spacing between the linear regions 8b extending in the direction perpendicular to the specific direction do not necessarily have to be the same and are often different from each other. . Therefore, in the normal wafer 2, the linear regions 8 (8a and 8
A plurality of rectangular areas 10 are divided by b). A required circuit pattern is applied to this rectangular area IO. In such a wave 2, by utilizing the above-mentioned orientation diagram If, either one of the linear regions 8a or 8b, in the illustrated case the linear region 81, is aligned in a predetermined reference direction, that is, in the X direction (see FIG. 1). ) is placed on the holding means 4 so that the angle of inclination is within a range of, for example, about ±1.5 degrees to about 3.00 degrees or less.

Continuing the explanation with reference to FIG. 1, the holding means 4, which may be of a known form per se, holds the wafer 2 placed on its surface sufficiently reliably by vacuum suction or the like. This holding means 4 is mounted movably in the X direction, the X direction, and the θ direction by an appropriate support mechanism (not shown). A moving means 12 is drivingly connected to the holding means 4 to move it from place to place with sufficient precision. In the illustrated embodiment, the moving means 12 includes an X-direction moving source 14
, an X-direction movement source 16, and a θ-direction movement source 18. The X-direction displacement source 14, which advantageously consists of a pulse motor, when actuated causes the holding means 4 to be moved a required distance in the X-direction with an accuracy of, for example, 1 μm. The X-direction movement source 16, which advantageously comprises a pulse motor, when activated, moves the holding means 4 a required distance in the X-direction, that is, in a direction perpendicular to said X-direction, with an accuracy of, for example, 1 μm. .

The theta displacement source 18, which likewise advantageously consists of a pulse motor, when actuated causes the holding means 4 to
is moved by a required angle in the θ direction with zero accuracy, for example, about 0.0015 degrees, that is, rotated about the central axis 20 of the holding means 4.

The illustrated semiconductor wafer cutting apparatus includes a rotating blade 22 which is preferably formed from fixed diamond abrasive grains.
is provided. The rotary blade 22 constituting the wafer cutting means is rotatably mounted around a central axis 24 substantially parallel to the X direction and movable in the X direction, and is driven by an appropriate drive such as an AC motor. and a suitable drive source (not shown) such as a DC motor.
It is caused to reciprocate in the X direction at the required speed.

In the illustrated semiconductor wafer cutting apparatus, the holding means 4
The wafer 2 is placed on the holding means 4 by the supply means (not shown) while the wafer 2 is present in the supply and discharge area at or near the position indicated by the solid line in FIG. Then, by finely adjusting the position of the holding means 4, as will be described in detail later, the wafer 2 held on the holding means 4 is sufficiently accurately positioned at a predetermined position with respect to the rotating blade 22. After that, the holding means 4
The holding means 4 and the wafer 2 held on its upper surface are positioned in a cutting start area adjacent to the rotary blade 22, as shown by the two-dot chain line in FIG. Next, a cutting movement is performed in which the rotary blade 22 is rotated and moved in the X direction so that the wafer 2 is subjected to the action of the rotary blade 220 that is rotationally driven, and a linear region 80 existing on the surface of the wafer 2 is performed.
The so-called index movement of moving the holding means 4 in the X direction by the distance d + w (FIG. 2) is carried out alternately, thus cutting the wafer 2 along the linear region sb (or 8 m) present on its surface. Next, the holding means 4 is moved 90 degrees in the θ direction about its central axis @ 20, and then the cutting movement and the indexing movement are performed alternately, thus moving the wafer 2 into a linear area existing on its surface. 8a (or 6b). After that, the holding means 4 is moved backward a predetermined distance in the X direction, and the holding means 4 is returned to the supply and discharge area. The wafer 2 is ejected from the holding means 4 by suitable ejection means (not shown), which may be of a known form per se, and the next wafer 2 is placed on the holding means 4 by said feeding means (not shown). The cutting of the wafer 2 by the rotating blade 22 is as follows:
As is well known to those skilled in the art, this is done not over the entire thickness of the wafer 2, but with only a small residual uncut thickness, so that the rectangular areas 10 (FIG. 2) are not completely separated. (In this case, the rectangular area 10 is completely separated by applying some force later to break the uncut portion, thus producing a chip.)

Alternatively, an adhesive tape may be attached to the back surface of the wafer 2 in advance so that the rectangular regions 10 are not separated into individual parts even if the wafer 2 is cut across its entire thickness. (The rectangular area 10 is then completely separated by peeling off the adhesive tape, thus producing a chip).

Continuing the explanation with further reference to FIG. 1, in the embodiment shown, the holding means 4 is located above the holding means 4 and the wafer 2 held on its surface when the holding means 4 is present in the supply and discharge areas. A stationary enlarging optical means 26 and an imaging means 28 optically connected to the enlarging optical means 26 are provided. The illustrated magnifying optical means 26 is composed of a binocular microscope having two light entrance openings 30a and 30b located at an appropriate interval, for example, about 40 m+, in the X direction. Therefore, images of two parts of the surface of the wafer 20 held on the holding means 4 at a predetermined interval t-t in the The image is enlarged and sent to the imaging means 28 as a split image. Magnifying optical means 2
The magnification by 6 may be, for example, about 20 times. If desired, a variable magnification optical means can be used which allows the magnification to be changed stepwise or continuously.

The imaging means 28 may be of any type as long as it is capable of outputting an analog signal indicating the density of the x-y matrix array pixels according to the captured image, but is particularly suitable for solid-state cameras! Preferably, it consists of a solid-state camera having a plurality of imaging elements arranged in a -7 matrix, such as a CCD, CPD or MOS. In the illustrated example, the imaging means 28 has a 256×25
It consists of a solid-state camera with six COD'i arranged in a matrix. The 128 x 256 CODs located in the left half of the 256 x 256 CODs receive the image incident on the winding side human light aperture 30m of the enlarging optical means 26, and the remaining CODs located in the right half. 128X25
The six CODs have the right entrance aperture 3 of the magnifying optical means 26.
An image of light incident on 0b is input. Each of the 256x256 CCDs has a density (gr) of the image input to it.
output an analog signal having a voltage according to ay 1evel). The solid-state camera with 256 x 256 CODs is equipped with automatic gain adjustment means (not shown), known per se, for automatically adjusting the gain of the output analog signal according to the actual density of the captured image. It is convenient if it is attached or built-in.

Referring to FIG. 3, which is a block diagram schematically showing various electronic and electrical means provided in connection with the imaging means 28, the output analog signal of the imaging means 28 is VD (analog - Digital) converting means 32, which converts the input analog signal into a multi-value digital signal, which may be, for example, 8 bits (therefore, 2g=256 steps). The multivalued digital signal is then sent to and stored in the image frame memory 34. Image frame memory 3 in the illustrated example
4 has a storage capacity of at least 256 x 256 x 8 pits, and therefore the 25
256×25, each corresponding to the density of 6×256 pixels
It is composed of a RAM that can store six 8-bit multivalued digital signals.

In the illustrated example, the image frame memory 34
A central processing unit (CPU) 36', which can be a microprocessor with multiple built-in RAMs,
A non-pattern matching position sensing means, generally designated 38, is connected.

The illustrated non-pattern matching type position detection means 38 includes an arithmetic processing means 40 and a linear area detection means 42 . , the arithmetic processing means 4o includes a differential processing circuit 44, a threshold processing circuit 46, and a parity check processing circuit 48. The calculation processing means 4° performs the following operations according to the operator's selection.
Three modes: differential operation processing and threshold processing mode;
Simple thresholding.

The multivalued digital signal stored in the image frame memory 34 is subjected to arithmetic processing according to either the mode and the threshold processing or parity check processing mode.
Generate a value digital signal.

The binary digital signal generated by the arithmetic processing means 4o is stored in the RAM built in the central processing unit 36. The linear area detection means 42 detects the linear area 8a (or s
b) and thus detect the relative position of the linear region 8a (or sb). The non-pattern matching type position detecting means 38 including the arithmetic processing means 40 and the linear area detecting means 42 is
The details of the structure and operation and effect of the non-pattern matching position detection means 38 may be in the form described in detail in the specification and drawings of Patent Application No. 162031 of 1982. Patent application No. 16203
The details will be referred to the specification and drawings of No. 1, and the explanation will be omitted in this specification.

In the illustrated embodiment, display means 50 are also provided, which advantageously consist of a cathode ray tube (CRT). This display means 50 displays the multivalued digital signal outputted by the percentage conversion means 32 and stored in the RAM built in the central processing unit 36 in response to manual operation of a switching means (not shown). An image corresponding to the binary digital signal or a signal stored in a key pattern memory to be described later is selectively displayed visually. In the illustrated display means 50, an image related to an image inputted to the imaging means 28 from the left light entrance aperture 3Qa of the enlarging optical means 26 is displayed in the left half thereof, and an image related to the image inputted to the imaging means 28 is displayed in the right half thereof. 26 from the right light entrance aperture 30b of the imaging means 28
The images related to the image input to the image display screen are respectively enlarged and displayed at a total magnification of about 260 times, for example.

A key pattern memory 52 and a pattern matching type position detection means 54 are also connected to the image frame memory 34 via the central processing unit 36 .

Key pattern memory 5 that can be configured from RAM
2 stores a signal indicating the position of a key pattern and a key pattern in a specific area on the surface of the wafer 2 when the wafer 2 held on the holding means 4 is in a predetermined position. An example of a method for inputting signals to the key pattern memory 52 is as follows. When inputting a signal to the key pattern memory 52, the holding means 4
The sample wafer 2 is placed on top of the sample wafer 2, and then the X-direction movement source 14, the X-direction movement source 16, and the θ-direction movement source 18 are manually activated as appropriate to move the holding means 4, and the sample wafer 2 on the holding means 4 is moved. 2 to the magnifying optical means 2
8, manually position it at the required position. During such manual positioning, the display means 50 is made to visually display the multivalued digital signal outputted by the percentage conversion means 32, and the image displayed on the display means 50 is observed. As schematically shown in FIG. 4, the sample is arranged so that the stop line of the linear region 8a on the surface of the sample wafer 2 substantially coincides with the lateral center line, that is, X-Xl on the display screen of the display means 50. wafer 2
position. Next, the cursor 55 is manually positioned in the specific areas 56L and 56R on the left half and right half of the display screen of the display means 50, respectively. The cursor 55, and thus the specific areas 56L and 56R specified by the cursor 55, may be squares having dimensions corresponding to, for example, 32×32 pixels (and therefore corresponding to 32×32 CCDs in the imaging means 28). Specific areas 56L and 56R specified by cursor 55
is preferably a region having a remarkable feature, for example, a region at the intersection of the linear region 8a and the linear region 8b. After that, the specific area 5 of the multivalued digital signal stored in the image frame memory 34 is
32x32-, ., present in 6L and 56R. 1024
A signal corresponding to each pixel is sent to the key pattern memory 52 and stored therein. At the same time, signals indicating the positions (i.e.,
and store it. Thus, key pattern memory 5
2 stores a multivalued digital signal indicating the pattern of the specific areas 56L and 56R, that is, a key pattern, and also stores X-coordinate and y-coordinate signals indicating the position of the key pattern.

In a preferred embodiment, a secondary key pattern storage operation is performed after the primary key pattern/storage operation described above.

In this sub-key pattern storage operation, the cursor 55 is moved to the specific areas 56L and
It is manually positioned in an appropriate area separate from 6R, ie, MJ, sub-specific areas 58L and 58R. After that, in the same manner as in the key pattern storage operation described above, the sub-specific area 58L is
and a multi-fI indicating the 58R pattern, that is, the subkey pattern.
The I4 digital signal is stored in the key pattern memory 52. Then, the X-coordinate and y-coordinate signals indicating the position of the sub-key pattern are still stored in the key pattern memory 52.

When the above-described key pattern storage operation and sub-key pattern storage operation are completed, the θ direction movement source 18 is manually operated to rotate the sample wafer 2 held on the holding means 4 by 90 degrees.

Next, the X-direction movement source 14 and the X-direction movement source 16 are operated manually as necessary, so that the center line of the linear region 8b on the surface of the sample wafer 20 is aligned as schematically shown in FIG. The sample wafer 2 is positioned so as to match the horizontal center line, that is, X-XIM on the display screen of the display means 50. Then, a key pattern storage operation similar to the key pattern storage operation described above is performed. That is, the cursor 55 is manually positioned in the specific areas 60L and 60R on each of the left and right sides of the display screen of the display means 5o, and then a polygon indicating the pattern of the specific areas 60L and 60H, that is, the key pattern is displayed. Key pattern memory 52 for value digital signal
In addition, X and Y coordinate signals indicating the position of the key pattern are stored in the key pattern memory 52. In addition, after the sample wafer has been rotated by -2'i90 degrees, the central processing unit uses a signal indicating the amount of movement in the X direction and the y direction performed to position the sample wafer 1 in the state shown in FIG. 5 as a rotational displacement signal. 36 (or key pattern memory 52)
Remember.

In a preferred embodiment, further, after the key pattern storage operation, a secondary key pattern storage operation similar to the secondary key pattern storage operation described above is performed.

That is, in each of the left half and right half of the display screen of the display means 50, the cursor 55 is manually moved to an appropriate area separate from the temporary areas 60L and 60R, that is, the sub-specific areas 62L and 62R. Then, the pattern of the sub-specific areas 62L and 62R, that is, the multi-level digital signal indicating the sub-key pattern is stored in the key pattern memory 52, and the X-coordinate and y-coordinate signals indicating the position of the sub-key pattern are stored as the key pattern. It is stored in the memory 52.

The pattern matching type position detecting means 54 detects an image taken by the imaging means 26 on the surface of the wafer 2 held on the holding means 4 and to be automatically positioned at a desired position, and therefore displayed on the display means 50. The same pattern as the key pattern or sub-key pattern is detected in the image, and thus the relative position of the linear area 8a or 8b is detected. An example of such pattern matching type position detection will be explained as follows. The same pattern as the key pattern of the specific area 56L is displayed in the image inputted to the imaging means 28 from the left side light entrance aperture 30m of the enlarging optical means 26, and therefore displayed on the left half of the display screen of the display means 50. The case of detection will be explained with reference to the flowchart shown in FIG. 6. First, in step n-1, the cursor 55 is positioned at a predetermined position, for example, the upper left corner of the display screen of the display means 50. A matching area to match the key pattern is defined. Next, the process proceeds to step n-2, where the degree of matching P between the matching area and the key pattern is calculated. The calculation P of the matching degree is calculated by the key pattern memory 52.
A multi-value digital signal indicating a key pattern, that is, a 32 x 32 multi-value digital signal indicating the density of 32 x 32 pixels in the specific area 56L, and a multi-value digital signal indicating the density of 32 x 32 pixels in the matching area, which are stored in A multivalued digital signal 46 indicating the density is input from the imaging means 28 to the image frame memory 34 via the % conversion means 32.
It can be calculated based on 32×32 multivalued digital signals in the signal. The matching degree P itself can be calculated using the following formula A, for example, where f is a value corresponding to the density of each of the 1.6 x 16 pixels in the matching area, and i is the average value of f.
), g is a value corresponding to the density of each of the 16×16 pixels in the key pattern, i is the average value of g, (t
, j) indicates the row and column of each pixel, and can therefore be calculated based on (l=1 to 16.J=1 to 16). In calculating the matching degree P based on the above formula A, the density deviation value of each pixel in the matching area (that is, the value obtained by subtracting the average density value from the actual density value) and each pixel in the key pattern are calculated. The difference from the density deviation value is added, and therefore, the so-called density gain fluctuations caused by fluctuations in illuminance for the verification area are eliminated.
A sufficiently reliable matching degree P is obtained.

In order to simplify the performance-processing, C f in the above formula A
(1,j)-j], and Cg(t,j)-i-1
Formula 1 below, which adds binarization processing to the names of % Here, U means binarization operation, and X>.

If U(X)=1. When x≦0, U(X)=O. The matching degree P can also be determined based on the following.

In order to further increase the reliability of the required matching degree P, based on the so-called normalized correlation, i.e., the above formula B. Here, 'l'l glg and (i, j)
The matching degree P can also be determined based on , which is the same as in the case of .

Therefore, when calculating the matching degree P based on the above formula A, B or C, all pixels (32×32
= 1024), in order to speed up the calculation speed, a plurality of specific pixels among the pixels in the matching area, for example, 32 specific pixels selected one in each row name column, are used. It is also possible to perform correlation processing only on the data. In particular, based on the above formula C, the matching degree P
When calculating , it was confirmed that sufficiently good results could be obtained for #Okatonde semiconductor wafers even if correlation processing was performed only on multiple specific pixels among the pixels in the matching area. has been done.

Following the calculation of the matching degree P, in step n-3, it is determined whether the calculated matching degree P is greater than or equal to a predetermined value. The predetermined threshold value is set by the operator as appropriate (for example, by trial and error method) and stored in the key pattern memory 52.
Or it can be stored in the RAM in the central processing unit 36. If the calculated degree of matching P is not equal to or higher than the predetermined darkness value, that is, if the degree of matching is relatively low, the process proceeds to step n-4, and the matching degree P is inputted to the imaging means 28 from the left light entrance aperture 30a of the enlarging optical means 26. It is determined whether the cursor 55 has been moved over the entire area of the image, that is, the image displayed on the left half of the display screen of the display means 50. If the movement of the cursor 55 over the entire area has not been completed, the process proceeds to step n-5, where the cursor 55 is moved by one pixel in the x direction and/or y direction and moved to the next matching area. Thereafter, the matching degree P is calculated in step n-2, and it is determined in step n-3 whether the calculated matching degree P is greater than or equal to a predetermined threshold value. If the calculated degree of matching P is greater than or equal to the predetermined threshold, that is, if the degree of matching is relatively high, the process proceeds from step n-3 to step n-6, and the position of the matching area and the degree of matching P are determined by the pattern matching method position. The data is stored and restored in the RAM built in the detection means 54 or the RAM built in the central processing unit 36. Then, the above step n
-Proceed to 4. In this way, the enlarging optical means 26 adjusts the matching degree P over the entire area of the image input to the imaging means 28 from the left light entrance aperture 30 &, therefore, the image displayed on the left hand portion of the display screen of the display means 50. When the calculation of the matching degree P and the determination as to whether or not the calculated matching degree P is equal to or higher than the predetermined threshold are completed, the process proceeds from step n-4 to step n-7, and the matching degree P restored in the step n-5 is calculated. The largest one is selected, and it is determined that the matching area having the maximum matching degree Pn'IILx is the same as the key pattern, that is, the specific area 56L. In this way, from the position (X and y coordinate position) of the matching area having the maximum matching degree Pmx and the position (X and y coordinate position) of the key pattern, that is, the specific area 56L, it is inevitable that the relative position of the linear area 8a position # is detected.

From the left light entrance aperture 3°a of the enlarging optical means 26, the imaging means 2
In the image manually inputted in 8, and therefore displayed on the left half of the display screen of the display means 50, the sub-specific area 5
When detecting the same pattern as the sub-key pattern of 8L, the key pattern of the specific area 60L, or the sub-key pattern of the sub-specific area 62L, or the enlarging optical means 26
In the image inputted to the imaging means 28 from the right light entrance aperture 30b, and therefore displayed on the right half of the display screen of the display means 50, the key pattern of the specific area 56 and the sub-specific area 58H are displayed. Even when detecting the same pattern as the key pattern, the key pattern of the specific area 60H, or the sub-key pattern of the sub-specific area 62H, the sixth
A procedure similar to the procedure described above with reference to the flowchart shown in the figure may be performed.

In an automatic precision alignment system constructed according to the present invention, the moving means 12, more specifically, the X-direction movement source 14, the Y-direction movement source 16, and the θ-direction movement source 18 are provided.
A movement control means 64 is provided for controlling the operation of the wafer 2 and positioning the wafer 2 held on the holding means 4 at a desired position.

The movement control means 64 controls the linear area 8a (or gb) by the non-pattern matching type position detection means 38.
The moving means 12 is actuated in response to the detection of the relative position of the wafer 2, thus performing the primary positioning of the wafer 2, and then determining the relative position of the linear area 8& (and Bb) by the pattern matching type position detection means 54. It is important to actuate the moving means 12 in response to the detection of the wafer 2, thus performing a secondary positioning of the wafer 2.

7-A to 7-C show an example of a flowchart for positioning by the moving means 64.

Referring to FIG. 7-A, in step ro-1, two images input to the imaging means 28,
That is, either one of the image inputted to the imaging means 28 from the left light entrance aperture 30a of the enlargement optical means 26 or the image inputted to the imaging means 28 from the right light entrance aperture 30b of the enlargement optical means 26, for example, the enlargement optical means 26 left light entrance aperture 30
It is determined whether the non-pattern matching type position detection means 38 can detect the linear region 8a in the image inputted to the imaging means 28 from point a. Holding means 4
The linear region 8a does not exist in the image due to an error in the placement position of the upward way...2, or the linear region 8a is localized in the captured image due to printing defects, etc. If the non-pattern matching type position detecting means 38 is unable to detect the linear area 8a, the process proceeds to step m-2, where the central processing unit 36 determines the amount of y-direction movement wJ. Set. This amount of movement in the y direction is an appropriate value that is also smaller than the interval d between the linear regions 8a.
For example, it may be about 300 μm. Next, the process proceeds to step m-3, where the amount of y-direction movement is added to the current y-direction position of the holding means 4 and the wafer 2 held therein. Next, the process proceeds to step m-4, where the added movement amount in the y direction and the interval d are compared.

Then, when the interval d is larger (if the y-direction movement amount is further added after the y-direction movement is returned multiple lines, the interval d becomes smaller), proceed to step m-5.
The control means 64 operates the y-direction movement source 16 to move the holding means 4 and the wafer 2 held therein in the y-direction by the above-mentioned y-direction movement amount. Then, the process returns to step m-1.

In step m-4 above, when the interval d is smaller,
That is, by adding the amount of movement in the y direction multiple times, the distance d is smaller than the added amount of movement in the y direction. If the total amount of movement in the y direction exceeds the above interval d, step m
-6, increments the X-direction movement counter built in the central processing unit by 1, and then proceeds to step m-7.
Then, it is determined whether the count value of the X-direction movement counter has reached 4 (that is, it is determined whether the X-direction movement has already been repeated three times). If the count value of the X-direction movement counter is not 4, the process proceeds to step m-8, in which the movement control means 64 activates the X-direction movement source 14 to cause the holding means 4 and the holding means 4 to hold. The wafer 2 is moved in the X direction by a predetermined amount, which may be about 170 μm, for example. Then, proceed to step m-9,
The y-direction movement is set to be reversed (that is, if the y-direction movement was previously performed in the positive direction, it is now set to be performed in the negative direction). And the above step m
-Proceed to 3. Therefore, in the illustrated example, if the linear region 8a cannot be detected, the holding means 4 and the wafer 2 held therein are It will be understood that the detection of the linear region 8a is repeated by moving in a so-called zigzag manner in the X direction and the X direction (up to three times in the X direction).

In the above step m-1, when the non-pattern matching type position detection means 38 detects the linear region 8a, the process proceeds to step m-10, and the center in the y direction of the image being imaged (i.e., It is determined whether or not there is a deviation Pl between the center of the linear region 8m and the center of the linear region 8m.

Then, if the above deviation exists, step +n-
11, the movement control means 64 activates the X-direction movement source 16 to move the holding means 4 and the wafer 2 attached thereto in the Y direction by the above-mentioned deviation, and thus the linear region 8a is imaged. Position it at the center of the image in the y direction. Thereafter, the process proceeds to step m-12, in which non-pattern matching is performed in the other of the two images input to the imaging means 28, in other words, in both of the two images input to the imaging means 28. Expression detection means 3
8 detects the linear region 8a. The linear region 8a is -, #! (Fig. 4), there is no linear region 81 in the other of the two images, and therefore the non-pattern matching position The detection means 38 is
If it is not possible to detect the linear region 8a in both images, the process proceeds to step m-13.
In the linear region 8a detected in one of the individual images,
Detect the inclination angle with respect to the X direction. To detect such an inclination angle, the holding means 4 and the wafer 2 held therein are moved by a predetermined distance in the X direction at least once, and the position of the linear region 8a in the Y direction before movement and the straight line after movement are determined. It can be conveniently detected based on the difference from the y-direction position of the shaped region 8a. In the case of the specific example shown,
A three-point check method is adopted to check the y-direction position of the linear region 8a in the image before movement and the X-direction drive source 14.
After activating the holding means 4 and the wafer 2 held thereon by a predetermined distance of about 2 wm in the X direction, the position of the linear region 8a in the y direction, and further activating the X direction drive source 14. Based on the difference in the three y-direction positions from the y-direction position of the linear area 8a after the holding means 4 and the nine wafers 2 held therein are further moved the predetermined distance in the X direction, Detecting the tilt angle. Once the tilt angle is detected in this way, the process proceeds to step m-14, where rough alignment in the θ direction is performed according to the detected tilt angle. That is, in accordance with the detected inclination angle, the movement control means 64 drives the θ direction movement source 18 to move the holding means 4 and the wafer 2 held therein in the θ direction, that is, along the central axis 20 (FIG. 1). ), whereby the linear region 8a is rotated around X-
The X-rays (Fig. 4) and the rays are made parallel. When such coarse alignment in the θ direction is performed, as is easily understood, the linear region 8a will exist in both of the two images input to the imaging means 28. In step m-12, if the non-pattern matching type position detection means 38 detects the linear region 8a in both of the two images, the process proceeds to step m-15. In step m-15, it is determined which of two alignment procedures is selected, which can be selected by the operator as appropriate, depending on the characteristics of the wafer 2 to be aligned, for example. That is, should the primary positioning based on the detection of the relative position of the linear region 8a by the non-pattern matching type position detection means 38 be further continued, or should the relative position of the linear region 81 be detected by the non-pattern matching type position detection means 38? After completing the primary positioning based on position detection, the linear area 8 is detected by the pattern matching type position detection means 54.
It is determined whether to proceed to secondary positioning based on the detection of the relative positions of a and 8b. If the former is selected, the process proceeds to step m-16, and the y of the linear region 8a detected by the non-pattern matching type position detection means 3B in each of the above two images inputted to the imaging means 28 is It is determined whether the directional positions match. As is easily understood, if the linear region 8a is not parallel to the X direction with sufficient precision, the linear region 8a detected in each of two images separated at a predetermined interval in the X direction. There is a difference in the position of 8a in the y direction. If such a difference exists, the process proceeds to step m-17, where sufficiently precise θ alignment is performed. That is, depending on the above-mentioned difference, the movement control means 64 operates the θ direction movement source 18 to move the holding means 4 and the wafer 2 held therein in the θ direction, so that the linear region 8a is sufficiently precisely X
parallel to the direction. After that step m-1
Proceeding to step 8, the central processing unit 36 determines whether there is a shift or deviation between the center of the image being captured in the y direction (i.e., the line X-X shown in FIG. 4) and the center of the linear region 8a. Decide whether or not. If the deviation exists, the process proceeds to step m-19, where the movement control means 64 operates the y-direction movement source 16 to move the holding means 4 and the wafer 2 held therein by the amount of the deviation in the y-direction. In this way, the linear region 8a is positioned sufficiently precisely at the center of the image being captured in the y direction. Thereafter, the process proceeds to step m-20, in which the movement control means 64 operates the y-direction movement source 16 to move the holding means 4 and the wafer 2 held therein at a distance d between the linear regions 8a (FIG. 2). ) in the y direction. Next, the process proceeds to step m-21, where it is checked again whether or not the non-pattern matching type position detection means 38 detects the linear region 8a in both of the two images input to the imaging means 28. . Holding means 4 and wafer 2 held therein
Since the amount of movement in the y direction is equal to the distance d between the linear regions 8&, normally, the non-pattern matching type position detection means 38 detects the linear regions 8& in both of the above two images. However, due to printing defects etc., the above
If it cannot be confirmed that the non-pattern matching type position detection means 38 detects the linear region 8a in both images, the process proceeds to step m-22, and the movement control means 52 14 to move the holding means 4 and the wafer 2 f held thereon by a predetermined distance, for example, about 200 μm, in the X direction, and then proceed to step m-23, where both of the above two images It is checked again whether the non-pattern matching type position detection means 38 detects the linear region 8a. In step m-21 or -23, the non-pattern matching position detection means 3 is detected in both of the two images.
8 detects the linear region 8a, the process proceeds to step m-24 (FIG. 7-B), and thereafter, the linear region 8 is detected by the pattern matching type position detection means 54.
A secondary alignment based on detection of the relative positions of a and 8b is performed.

To explain with reference to FIG. 7-B, in step m-24, in one (or both) of the two images input to the imaging means 28, for example, from the left side light entrance aperture 30m of the enlarging optical means 26, In the image input to the imaging means 28, it is determined whether the pattern matching type position detection means 54 detects a key pattern, that is, the same pattern as the pattern of the specific area 56L (FIG. 4). When the pattern matching type position detection means 54 does not detect the same pattern as the key pattern, the process proceeds to step m-25, where the central processing unit 36 sets the amount of movement in the X direction. The amount of movement in the X direction may be an appropriate value that is smaller than the distance d between the linear regions 8b, for example, about 300 μm. Next, the process proceeds to step m-26, where the amount of X-direction movement is added to the current X-direction position of the holding means 4 and the way/.2 held therein. Next, the process proceeds to step m-27, where the added movement amount in the X direction and the interval d are compared.

Then, when the interval d is larger (if the X-direction movement amount is further added after the X-direction movement is repeated multiple times, the interval d becomes smaller), the process proceeds to step m-28. , the movement control means 64 operates the X-direction movement block 14 to move the holding means 4 and the wafer 2 held therein in the X direction.
It is moved in the X direction by the amount of directional movement. Then, the process returns to step m-24. In step m-27, when the distance d is smaller, that is, by adding the X-direction movement amount multiple times, the distance d is smaller than the added movement amount in the X-direction. 4 and the way/・2 held by it are moved in the X direction by the amount of movement in the X direction, the total amount of movement in the X direction is the distance d.
When it exceeds ', the process proceeds to step m-29. In this step m-29, the X-direction movement counter built in the central processing unit is incremented by 1, and then the process proceeds to step m-30 to check whether the count value of the X-direction movement amount counter is 2 or not. (In other words, it is determined whether the movement in the X direction has already been carried out once.) And X
If the count value of the directional movement force/force is not 4, the process proceeds to step m-31, where the small control means 64 activates the X direction movement source 16, and the holding means 4 and the holding means 4 are held therein. The wafer 2 is moved in the y direction by the distance d between the linear regions 8a. Next, the process proceeds to step m-32, where the X direction movement is set to reverse (that is, until then
If the direction movement was performed in the positive direction, it is set to be performed in the negative direction from now on). After that, the process returns to step m-24. Because it depends on this,
If the pattern matching type position detection means 54 does not detect the same pattern as the key pattern in step m-24, the holding means 4 and the The held wafer 2 is moved in a so-called zigzag pattern in the X direction and the y direction (movement in the y direction is performed only once), and detection of the same pattern as the key pattern is repeatedly performed.

Accordingly, from the above step m-23 to the above step m-2
4, the non-pattern matching type position detection means 38 has already completed sufficiently precise primary positioning based on the detection of the relative position of the linear area 8a. Except for extremely rare cases such as erroneous detection by the position detection means 38, the above step m-2
4, the pattern matching type position detection means 54
immediately detects a pattern that is the same as the key pattern. If the pattern matching type position detection means 54 detects the same pattern as the key pattern, the process proceeds to step m-33, where the center of the y-direction of the captured image (i.e., the x - X-ray) and linear region 8
It is determined whether there is a deviation or deviation from the center of a. If the above-mentioned deviation exists, the process proceeds to step m-34, where the movement control means 64 activates the y-direction movement source 16, thus moving the linear area 8a sufficiently to the center of the y-direction in the image being taken. Position precisely. Thereafter, the process proceeds to step m-35, where the movement control means 64 activates the X-direction movement source 14 to move the holding means 4 and the wafer 2 held therein in the X-direction by the distance d of the linear region 8b. move it to Next, the process proceeds to step m-36, in which the pattern matching type position detection means 54 detects a turn that is the same as the key pattern in one (or both) of the two images input to the imaging means 28. If it is confirmed that the pattern matching type position detection means 54 detects the same turn as the key pattern, step m-3
7, the center in the y direction of the image being captured (
That is, it is determined whether there is a shift or deviation between the line XX shown in FIG. 4 and the center of the linear region 8a. Then, if the above deviation exists, step m-
38, the movement control means 64 actuates the y-direction movement source 16, thus positioning the linear region 8a with sufficient precision at the center of the y-direction of the image being imaged.

Thereafter, the process proceeds to step m-39, in which it is determined whether the holding means 4 and the wafer/S2 held therein have been rotated by 90 degrees. If the 90 degree rotation has not been completed yet, proceed to step m-40;
The movement control means 64 operates the θ direction drive source 1B,
The holding means 4 and the wafer 2 held therein are rotated 90 degrees. Next, the process proceeds to step m-41, in which the movement control means 64 controls the X-direction drive source 14 and the Y-direction drive source 16.
is activated to rotate the rotational displacement signal stored in RI%yL (or in the key pattern memory 52) built in the central processing unit 36, that is, the sample wave at the time of key pattern storage operation, by 90 degrees. The holding means 4 and the wafer 2 held therein are moved in the X and y directions by an amount of movement corresponding to the amount of movement in the X and y directions after that. Thus, the same key patterns as the key patterns of the specific areas 60L and 60R (and the sub-key patterns of the sub-specific areas 62L and 62R) shown in FIG. 5 are present in the image input to the imaging means 28. ensure that After that, the process returns to step m-24. After returning to step m-24 via steps m-40 and m-41, in steps m-24 and m-36, the two images input to the imaging means 28 are In one (or both) of the above, the pattern matching type position detection means 54 detects the specific area 56.
Not L (or 56R) but the specific area 60L (60
It is determined whether the same pattern as the key pattern of R) is detected.

In accordance with the above procedure, secondary indexing and secondary positioning are performed, and thus the wafer 2 is positioned sufficiently quickly and precisely enough to substantially eliminate mispositioning.

Next, in step m-15, the primary positioning based on the detection of the relative position of the linear region 8a by the non-pattern matching type position detection means 38 is completed, and the linear region 8a is detected by the pattern matching type position detection means 54. A case will be described in which it is determined that it is necessary to shift to secondary positioning based on the detection of the relative positions of 8a and 8b. In this case, the process proceeds from step m-15 to step m-42 (FIG. 7-0). To explain with reference to FIG. 7-0, in step m-42, one of the two images input to the imaging means 28, for example, the magnifying optical means 26
It is determined whether or not the pattern matching type position detection means 54 detects the same pattern as the key pattern of the specific area 56L (FIG. 4) in the image input to the imaging means 28 from the left side light entrance aperture 30m. be done.

In step m-42, if the pattern matching type position detection means 54 does not detect the same pattern as the key pattern, the start of secondary positioning is stopped and the process returns to primary positioning again.

That is, the process proceeds to step m-43, where the movement control means 64 activates the X-direction drive source 14 to move the holding means 4 and the wafer 2 held therein by a predetermined distance α in the X direction. The predetermined amount α may be, for example, about 170 μm. Next, proceeding to step m-44, the imaging means 28
It is determined whether or not the non-pattern matching type position detecting means 38 detects the linear region 8& in both of the two images inputted. In step m-44, if the non-pattern matching type position detection means 38 detects the linear area 8ai in both of the two images, the process returns to step m-42. On the other hand, in step m-44, if the non-pattern matching type position detection means 38 does not detect the linear region 8a in both of the two images, the process proceeds to step m-45, and the movement control means 64
activating the directional drive source 14 to move the holding means 4 and the wafer 2 held therein by a predetermined amount -α in the X direction;
That is, it returns to the position before being moved by the predetermined amount α in step m-43. Thereafter, the process proceeds to step m-51, which will be described later.

In step m-42, if the pattern matching type position detection means 54 detects the same pattern as the key pattern, the process proceeds to step m-46, and the center in the y direction (i.e., the fourth It is determined whether there is a shift or deviation between the line XX shown in the figure) and the center of the linear region 8a. If the above-mentioned deviation exists, the process proceeds to step m-47, where the movement control means 64 activates the y-direction movement source 16, and the center of the y-direction of the image of the linear region 8a is then position with sufficient precision. After that, step m
-48, the movement control means 64 moves the X direction movement source 14
Then, the X-direction movement source 16 is activated to move the holding means 4 and the wobble 2 held therein to the specific area 56L.
The distance β and y in the X direction between the key turn (Fig. 4) and the sub key turn of the sub-specific area 58L (Fig. 4)
It is moved in the X direction and the X direction by a directional distance γ. Next, the process proceeds to step m-49, in which the turn matching type position detection means 54 detects the sub-specific region 58L (FIG. 4) in one of the two images inputted to the imaging means 28.
It is determined whether or not a no-turn that is the same as the sub-key pattern of is detected.

In step m-49, if the turn matching type position detection means 54 does not detect the same pattern as the subkey turn, the secondary position search is stopped and the process returns to the primary position. Go back, i.e. step m-50
The movement control means 64 moves the X direction drive source 14 and the
The directional drive source 16 is activated to move the holding means 4 and the wafer 2 held thereon by a predetermined distance -β in the X direction and by a predetermined distance -γ in the X direction, thus Return to the position before the direction movement and X direction movement. Next, the process proceeds to step m-51, where the movement control means 64 operates the X-direction drive source 16 to move the holding means 4 and the wafer 2 held therein by the distance d of the linear region 8a in the X direction. . Thereafter, the process proceeds to step m-52, in which it is determined whether the non-pattern matching type position detection means 38 detects the linear region 8a in both of the two images input to the imaging means 28. In step m-52, the non-pattern matching type position detection means 3 is detected in both of the two images.
If the linear region 8 is detected, the process returns to step m-42. On the other hand, in step m-52, 2
If the non-pattern matching type position detection means 38 does not detect the linear region 8a in both images, the process proceeds to step m-53, where the movement control means 64 activates the X-direction drive source 14, The holding means 4 and the wafer 2 held therein are moved in the X direction by a predetermined amount, for example, by about 170 μm. Thereafter, step m-54
Then, the non-pattern matching type position detection means 38 is applied to both of the two images that are inputted to the imaging means 28 again.
It is determined whether or not the linear region 8at is detected. In step m-54, the non-pattern matching type position detection means 38 detects the linear region 81 in both of the two images.
If detected, the process returns to step m-42.

In step m-49, if the pattern matching type position detecting means 54 detects the same pattern as the sub-key pattern, the process proceeds to step m-55, and the process proceeds to step m-55. It is checked whether there is a shift or deviation between x-x,@ ) shown in FIG. 4 and the center of the linear region 8a. If the deviation exists, the process proceeds to step m-56, where the movement control means 64 activates the X-direction movement source 16,
In this way, the linear region 8a is positioned with sufficient precision at the center of the image being captured in the X direction. Thereafter, the process proceeds to step m-57, and the step m-57 described above is also performed regarding the other of the two images input to the image pickup means 28, that is, the image input to the image pickup means 28 from the right light entrance aperture 30b of the enlarging optical means 26. -42 to m-56 are performed and it is determined whether they are friends or not. If steps m-42 to m-56 described above with respect to the other of the two images have not been performed, skip step m-42 above, and perform steps m-42 to m-56 described above with respect to the other of the two images. -56 is performed. On the other hand, if steps m-42 to m-56 described above have been performed with respect to the other of the two images, the process proceeds to step m-58, in which the pattern mat is Ching type position detection means 38
It is determined whether or not the detected positions of the linear regions 8a in the y direction match. If there is a difference in the position of the linear region 8a in the y direction detected in each of the two images, the process proceeds to step m-59, and the step m-
Sufficiently precise alignment in the θ direction is achieved in the same manner as in case No. 17.

Thereafter, the process proceeds to step m-60, where the center of the image being imaged in the y direction (i.e., X-X shown in FIG.
It is ascertained whether there is a deviation or deviation between the straight line 8a and the center of the linear region 8a. If the above-mentioned deviation exists, the process proceeds to step m-61, where the movement control means 64 activates the y-direction movement source 16, so that the linear region 8a is sufficiently moved to the center of the y-direction in the captured image. Position precisely. Thereafter, step m-62
The holding means 4 and the wafer 2 held therein proceed to
It is determined whether a 90 degree rotation of has already been performed. If the 90 degree rotation has not yet been completed, the process proceeds to step m-63, where the movement control means 64 operates the θ-direction drive source 18 to move the holding means 4 and the wafer 2 held therein. Rotate it 90 degrees. Then step m
-66, in the same manner as in step m-41 described above, the holding means 4 and the wafer 2 held therein are moved by the amount of movement in the X and Y directions based on the rotational displacement signal.
is moved in the X and Y directions. After that,
Return to step m-42 above. Therefore, the above step m
After returning to step m-42 via steps -63 and m-64, in step m-42, the specified area 6 is selected instead of the specified area 56L (also [56R).
It is determined whether or not the same pattern as the middle pattern of 0L (or 60R) is detected, and in step m-49, the sub-specific area 62L (or 62R) is detected instead of the sub-specific area 58L (or 58R). It is determined whether or not the same pattern as the secondary key pattern is detected.

In the manner described above, the secondary positioning following the secondary positioning is performed, and thus the wave 2 is positioned sufficiently quickly and with sufficient precision in the required position with virtually no mispositioning.

Although one specific example of the automatic precision alignment system configured according to the present invention has been described above in detail with reference to the isolated drawings, the present invention is not limited to such specific example, and does not depart from the scope of the present invention. It goes without saying that various modifications and modifications can be made without having to do so.

[Brief explanation of the drawing]

FIG. 1 is a simplified perspective view schematically showing a portion of a semiconductor wafer cutting apparatus equipped with an embodiment of an automatic precision alignment system constructed in accordance with the present invention. FIG. 2 is a partial plan view showing a portion of the surface of a typical wafer. FIG. 3 is a block diagram showing one specific example of an automatic precision alignment system constructed in accordance with the present invention. 4 and 5 are simplified diagrams illustrating designated positions of a specific area and a sub-specific area on a pull wafer on a display screen by a display means. FIG. 6 is a flowchart showing an example of a butter/matching type position detection procedure. FIG. 7-A, FIG. 7-B, and FIG. 7-C are 7p-charts showing an example of the alignment procedure. 2... Semiconductor wafer 4... Holding means 8 (8a and sb)... Linear region 12... Moving means 26... Enlarging optical means 28... Imaging means 32... Radiation conversion unit 34... Image frame memory 36... Central processing unit 38... Non-pattern matching type position detection means 40.
... Arithmetic processing means 42 ... Linear area detection means 50 ... Display means 52 ... Key batt memory 54 ... Butter/matching type position detection means 64 ...
・Movement control diagram 1

Claims (1)

[Claims] 1. There are a plurality of linear regions arranged in a grid on the surface, and a circuit pattern is applied to each of the plurality of rectangular regions partitioned by the linear regions. an automatic precision alignment system for positioning a semiconductor wafer in a desired position; a holding means for holding the semiconductor wafer; a moving means for moving the holding means; and a semiconductor wafer held by the holding means. Imaging means for capturing an image of at least a portion of the surface of the wafer and outputting an analog signal indicating the density of the X-)F matrix array pixels, and a signal corresponding to the analog signal output by the imaging means. an image frame memory for storing a key pattern corresponding to a specific area on the surface of the semiconductor wafer when the semiconductor wafer is in a predetermined position, and a key pattern memory for storing a signal indicating the position of the key pattern; Detecting a pattern identical to the key pattern in the image captured by the imaging means based on the signal stored in the image frame memory and the key pattern signal stored in the key pattern memory. a pattern matching type position detection means for detecting the relative position of the linear area; and a non-pattern matching type position detection means for detecting the relative position of the linear area by a method other than pattern matching. actuating the moving means in response to detection of the relative position of the linear region by the position detecting means and the non-pattern matching position detecting means, thus primary positioning of the semiconductor wafer held by the holding means; and then actuating the moving means in response to detection of the relative position of the linear area by the pattern matching position detection means, thus secondary positioning of the semiconductor wafer held by the holding means. An automatic precision positioning system characterized by comprising: a movement control means for performing the movement. 2. The key pattern memory is configured such that the semiconductor wafer is the first
at least one key pattern corresponding to at least one specific area on the surface when in a predetermined position;
at least one key pattern corresponding to at least one specific area on the surface when the semiconductor wafer is in a second position rotated 90 degrees with respect to the first predetermined position; In the secondary positioning, the control means
2. The automatic precision positioning system of claim 1, further comprising: performing positioning with respect to a predetermined position, then rotating said holding means 90 degrees, and then positioning with respect to said second predetermined position. 3. A patent characterized in that the image frame memory is equipped with a conversion means for converting the analog signal generated by the imaging means into a multi-value digital signal, and the image frame memory is characterized by the multi-value digital signal generated by the conversion means. An automatic precision alignment system according to claim 1 or 2. 4. The non-pattern matching position detection means includes arithmetic processing means for arithmetic processing the multivalued digital signal stored in the image frame memory to generate a binary digital signal; and linear area detection means for detecting the relative position of at least one side edge of the linear area based on the linear area, and thus detecting the relative position of the linear area. automatic precision alignment system. 56. According to claim 3 or 4, the signal indicating the key pattern stored in the key pattern memory is a multivalued digital signal corresponding to the density of a plurality of pixels in the key pattern. automatic precision alignment system. 6. The pattern matching type position detection means calculates the matching degree P by the following formula (%), where f is a plurality of pixels in the matching area of the surface of the semiconductor wafer held by the holding means. T is the average value of f, g is the value corresponding to the density of multiple pixels in the key pattern υ, and i is the average value of g. The automatic precision alignment system according to claim 5, which calculates based on , . 7. The pattern matching type position detection means calculates the matching degree P by the following formula %, where f is a plurality of pixels in the matching area on the surface of the semiconductor wafer held by the holding means. υ is the value corresponding to the density of each pixel in the key pattern, f is the average value of f, g is the value corresponding to the density of multiple pixels in the key pattern, and i is the average value of g. The automatic precision alignment system according to claim 5, wherein the automatic precision alignment system is calculated based on the following: C and U mean a binarization operation. 8. The pattern matching type position detection means calculates the matching degree P by the following formula, where f corresponds to the density of each of a plurality of pixels in the matching area of the surface of the semiconductor wafer held by the holding means. 7 is the average value of f, 1g is the value corresponding to the density of multiple pixels in the key pattern), and i is the average value of g. An automatic precision alignment system according to claim 5. 9. The imaging means is composed of a solid-state camera having a plurality of imaging elements arranged in an XF matrix;
An automatic precision alignment system according to any one of claims 1 to 8. 10. An enlarging optical means is disposed between the surface of the semiconductor wafer and the imaging means, and an image of at least a portion of the surface of the semiconductor wafer is enlarged by a predetermined magnification by the enlarging optical means. The automatic precision positioning system according to any one of claims 1 to 9, wherein the automatic precision positioning system is inputted to the imaging means. 11. The automatic precision alignment system of claim 10, wherein said imaging means images two mutually spaced apart portions of said surface on said semiconductor wafer. 12. The automatic precision alignment system of claim 11, wherein the two parts are spaced apart from each other in either the X direction or the Y direction. 13. In the next positioning, the movement control means:
13. The automatic precision positioning system according to claim 12, wherein the linear region is caused to exist in at least both of the two portions imaged by the imaging means.