Notch detection method

Abstract

The present application provides a notch detection method capable of correctly detecting a notch of a wafer. The notch detection method detects a notch formed on a peripheral portion of a wafer, and includes: a wafer placement step of placing the wafer on a rotary table; an image acquisition step of acquiring an image of the peripheral portion of the wafer; a profile data acquisition step of acquiring profile data including coordinates of a profile of the wafer based on the image; a virtual circle calculation step of calculating a virtual circle approximating the profile of the wafer based on the coordinates of the profile of the wafer; a special shape region determination step of determining whether a special shape region exists on the peripheral portion of the wafer based on a distance between the virtual circle and the profile of the wafer in a radial direction of the wafer; and a first notch determination step of determining whether the special shape region is a notch based on a distance between the virtual circle and a leading end of the special shape region in the radial direction of the wafer.

Description

Technical Field

[0001] This invention relates to a notch detection method for detecting notches that indicate the crystal orientation of a wafer. Background Technology

[0002] In the manufacturing process of device chips, a wafer in which devices are formed in multiple regions divided by a grid of spacers (predetermined dividing lines) arranged in a grid pattern is used. By dividing the wafer along the spacers, multiple device chips, each containing a device, can be obtained. These device chips are then assembled into various electronic devices such as mobile phones and personal computers.

[0003] In wafer dicing, for example, a cutting device is used. The cutting device rotates a ring-shaped cutting tool and cuts into the workpiece, thereby cutting the workpiece. On the other hand, wafer dicing is sometimes performed by laser processing. In laser processing of wafers, a laser processing apparatus is used, which has a laser irradiation unit that irradiates the workpiece with a laser beam.

[0004] Furthermore, in recent years, with the miniaturization of electronic devices, there is a demand for thinner device chips. Therefore, sometimes a wafer thinning process is performed before wafer dicing. In wafer thinning, grinding equipment is used, which uses grinding wheels with multiple grinding tools to grind the workpiece. The ground wafer is then diced along spacers, thereby manufacturing thin device chips.

[0005] Furthermore, a notch, indicating the crystal orientation of a silicon wafer, is formed on a crystalline wafer. The notch is formed by intentionally cutting away a portion of the outer periphery of the wafer. When processing the wafer using the various processing apparatuses described above, the notch is detected, and the wafer's orientation is adjusted according to the position of the notch. This allows processing to be performed with the wafer positioned in a predetermined orientation.

[0006] For example, Patent Document 1 discloses a grinding apparatus having an alignment unit for aligning wafers. The alignment unit adjusts the center position and orientation of the wafer based on the location of the notch formed on the outer periphery of the wafer by using a camera to photograph the wafer.

[0007] Patent Document 1: Japanese Patent Application Publication No. 2011-40637

[0008] When processing crystalline wafers using a processing apparatus, the processing characteristics of the wafer may vary depending on the crystal orientation of the wafer, depending on the processing requirements. Therefore, the processing apparatus detects the notches on the wafer and processes the wafer with the wafer positioned in a predetermined orientation according to the location of the notches.

[0009] However, sometimes defects other than notches are formed on the outer periphery of the wafer. For example, during machining processes such as cutting and grinding, defects may occur on the outer periphery of the wafer due to contact between the machining tool and the wafer, accidentally forming a notch-like cut. In such cases, when inspecting the notches on the wafer in subsequent processes, it is sometimes difficult to determine whether the cut on the outer periphery of the wafer is a notch or an accidental defect caused by machining, and the defect caused by machining may be mistakenly identified as a notch.

[0010] When a defect accidentally formed on the outer periphery of a wafer is mistakenly identified as a notch, the processing apparatus aligns the wafer and performs processing under the assumption that the defect is a notch. As a result, processing is performed with the wafer positioned in the wrong orientation, potentially leading to processing defects on the wafer. Summary of the Invention

[0011] The present invention was made in view of this problem, and its object is to provide a notch detection method that can accurately detect notches on wafers.

[0012] According to one aspect of the present invention, a notch detection method is provided to detect notches formed on the outer periphery of a disk-shaped wafer and showing the crystal orientation of the wafer. The notch detection method includes the following steps: a placement step, in which the wafer is placed on a rotating stage and a portion of the outer periphery of the wafer is positioned within the imaging area of ​​a camera; an imaging step, in which the rotating stage is rotated while the camera is used to take multiple images of the wafer, thereby acquiring an image of the outer periphery of the wafer; a contour data acquisition step, in which contour data, including the coordinates of the wafer's contour, is acquired based on the image; an imaginary circle calculation step, in which an imaginary circle approximating the wafer's contour is calculated based on the coordinates of the wafer's contour; an irregular shape region determination step, in which an irregular shape region is determined to exist on the outer periphery of the wafer based on the radial distance between the imaginary circle and the wafer's contour; and a first notch determination step, in which an irregular shape region is determined to be a notch based on the radial distance between the imaginary circle and the leading edge of the irregular shape region.

[0013] In addition, the preferred notch detection method further includes the following steps: a sharpness calculation step, which calculates the sharpness of the irregularly shaped area; and a second notch determination step, which determines whether the irregularly shaped area is a notch based on the sharpness of the irregularly shaped area.

[0014] In addition, the preferred notch detection method further includes the following steps: a segmentation step, dividing the image into a first image and a second image by dividing the irregularly shaped region into two parts along a straight line along the radial direction of the wafer; a first similarity calculation step, calculating a first similarity corresponding to the similarity between the first image and the flipped image, the flipped image being obtained by flipping the second image around the straight line as an axis; a second similarity calculation step, calculating a second similarity corresponding to the similarity between the first image and the second image; a third notch determination step, determining whether the irregularly shaped region is a notch based on the first similarity; and a fourth notch determination step, determining whether the irregularly shaped region is a notch based on the second similarity.

[0015] In one aspect of the notch detection method of the present invention, whether an irregularly shaped region formed on the outer periphery of a wafer is a notch is determined based on its size or shape. Therefore, even if a notch is formed on the outer periphery of the wafer, it is less likely to be mistakenly identified as a notch, and notches can be detected correctly. Attached Figure Description

[0016] Figure 1 This is a three-dimensional view of the chip.

[0017] Figure 2 This is a perspective view showing the notch detection device.

[0018] Figure 3 This is a flowchart illustrating the operation of the notch detection device.

[0019] Figure 4 (A) is an image diagram showing the first image representing the region where the notch is notched. Figure 4 (B) is an image diagram showing the second image depicting the area where the notch exists. Figure 4 Image (C) is a third image showing the region with a semi-circular defect. Figure 4 Image (D) is a fourth image showing the region with a fan-shaped defect. Figure 4 (E) is an image diagram showing the fifth image, which depicts a region with a rectangular defect.

[0020] Figure 5 (A) is an image diagram showing the first image with an imaginary circle. Figure 5 (B) is an image diagram showing the second image with an imaginary circle.

[0021] Figure 6 (A) is an image diagram showing the distance D1 in the irregularly shaped region of the second image. Figure 6(B) is an image diagram showing the distance D2 in the irregularly shaped region of the third image.

[0022] Figure 7 (A) is an image showing the front end of the irregularly shaped region in the second image. Figure 7 (B) is an image diagram showing the front end of the irregularly shaped region in the fourth image.

[0023] Figure 8 (A) is an image diagram showing the second image segmented by dividing lines. Figure 8 (B) is an image diagram showing the fifth image, which has been segmented by the dividing lines.

[0024] Figure 9 (A) is an image diagram showing the second image containing the flipped image. Figure 9 (B) is an image diagram showing the fifth image containing the flipped image.

[0025] Label Explanation

[0026] 11: Wafer; 11a: Front side; 11b: Back side; 11c: Notch (cut-out area); 13: Spacer (pre-defined dividing line); 15: Device; 2: Notch detection device (notch detection unit, notch detection section); 4: Rotary worktable (holding worktable); 6: Frame (main body); 6a: Upper surface; 6b: Recess (groove); 8: Holding component; 8a: Suction surface; 10: Spindle; 12: Rotation drive source; 14: Connecting part; 16: Flow path; 18: Suction source; 20: Camera (image capturing unit); 22: Image capturing area; 24: Control unit (control unit, control device); 26: Storage unit (memory); 30A~3 0E: Image; 32: Existing area (bright area); 34: Non-existent area (dark area); 36: Outline; 38A~38D: Irregularly shaped area; 38A1, 38A2, 38A2′: Irregularly shaped area; 38D1, 38D2, ​​38D2′: Irregularly shaped area; 40A~40D: Outline; 42: Imaginary circle; 44A, 44B: Front end; 46A, 46B: Dividing line; 48A, 48B: First image; 48Aa, 48Ba: Area; 50A, 50B: Second image; 50Aa, 50Ba: Area; 50A′, 50B′: Flipped image; 50Aa′, 50Ba′: Area. Detailed Implementation

[0027] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. First, a structural example of a wafer for detecting notches using the notch detection method of this embodiment will be described. Figure 1 This is a perspective view of chip 11.

[0028] For example, wafer 11 is a crystalline wafer formed from semiconductors such as silicon and shaped into a disk, having a front side 11a and a back side 11b that are substantially parallel to each other. Wafer 11 is divided into multiple rectangular regions by multiple spacer lines (predetermined dividing lines) 13 arranged in a lattice pattern in an intersecting manner. In addition, devices 15 such as IC (Integrated Circuit), LSI (Large Scale Integration), LED (Light Emitting Diode), and MEMS (Micro Electromechanical Systems) are formed in the multiple regions divided by the spacer lines 13 on the front side 11a of wafer 11.

[0029] A notch (cutout) 11c showing the crystal orientation of the wafer 11 is provided on the outer periphery of the wafer 11. The notch 11c is formed by cutting off a portion of the outer periphery of the wafer 11 from the outer periphery toward the center of the wafer 11. For example, the notch 11c is formed as a semi-circle when viewed from above.

[0030] The notch 11c is formed at a predetermined position corresponding to the crystal orientation of the wafer 11. Therefore, by confirming the position of the notch 11c, the crystal orientation of the wafer 11 can be determined.

[0031] Furthermore, there are no restrictions on the material, shape, structure, or size of the wafer 11. For example, the wafer 11 can be a disk-shaped wafer formed from semiconductors other than silicon (GaAs, InP, GaN, SiC, etc.), sapphire, etc. Additionally, there are no restrictions on the type, quantity, shape, structure, size, or arrangement of the devices 15.

[0032] Various processing techniques are performed on the wafer 11. For example, the wafer 11 is diced along the spacer 13, thereby manufacturing multiple device chips, each having a device 15. Alternatively, the wafer 11 is thinned before dicing, thereby obtaining thinner device chips.

[0033] In the dicing of wafer 11, for example, a cutting device or a laser processing device is used. The cutting device includes a chuck stage for holding wafer 11 and a cutting unit equipped with an annular cutting tool. The cutting tool is rotated and cuts into wafer 11, thereby cutting and dicing wafer 11. Similarly, the laser processing device includes a chuck stage for holding wafer 11 and a laser irradiation unit for irradiating wafer 11 with a laser beam. The laser irradiation unit irradiates wafer 11 with a laser beam, thereby performing laser processing on wafer 11.

[0034] In the thinning of wafer 11, for example, a grinding apparatus is used. The grinding apparatus has a chuck stage for holding wafer 11 and a grinding unit equipped with an annular grinding wheel containing multiple grinding tools. While rotating the grinding wheel, the grinding tools are brought into contact with wafer 11, thereby grinding and thinning wafer 11.

[0035] In addition, after the wafer 11 is processed, the processed wafer 11 is sometimes inspected. For example, the condition of processing marks (notches, damage, etc.) remaining on the wafer 11 is confirmed by an inspection device that observes the processed area (processed area) of the wafer 11.

[0036] During the processing or inspection of wafer 11, it is sometimes necessary to arrange wafer 11 in a specified orientation. Therefore, a notch detection device for detecting the notch 11c of wafer 11 is built into or connected to the various processing or inspection apparatuses described above. Furthermore, the orientation of wafer 11 is adjusted according to the position of the notch 11c detected by the notch detection device.

[0037] Figure 2 This is a perspective view showing the notch detection device (notch detection unit, notch detection section) 2. Additionally, in Figure 2 In this diagram, the X-axis (first horizontal direction) and the Y-axis (second horizontal direction) are perpendicular to each other. Additionally, the Z-axis (vertical, up-down, and height direction) is perpendicular to both the X-axis and Y-axis.

[0038] The notch inspection device 2 has a rotary stage (holding stage) 4 for holding the wafer 11. The upper surface of the rotary stage 4 is a flat surface that is approximately parallel to the X-axis and Y-axis directions, forming a holding surface 4a for holding the wafer 11.

[0039] Specifically, the rotary table 4 has a cylindrical frame (main body) 6 made of metal, ceramic, resin, or the like. A cylindrical recess (groove) 6b is formed in the center of the upper surface 6a of the frame 6. Furthermore, a disc-shaped holding member 8 for holding the wafer 11 is embedded in the recess 6b. The holding member 8 is a porous member made of a porous material such as porous ceramic, and contains a hole (flow path) connecting the upper surface of the holding member 8 to the lower surface.

[0040] The upper surface of the holding member 8 forms a circular attraction surface 8a that attracts and holds the central portion of the wafer 11. In addition, the upper surface 6a of the frame 6 and the attraction surface 8a of the holding member 8 are arranged on approximately the same plane, forming the holding surface 4a of the rotary table 4.

[0041] A cylindrical spindle 10, arranged along the Z-axis, is connected to the rotary table 4. The upper end of the spindle 10 is fixed to the lower end of the rotary table 4, and the lower end of the spindle 10 is connected to a rotary drive source 12. The rotary drive source 12 has a motor that rotates the spindle 10 about a rotation axis that is substantially parallel to the Z-axis. The rotary table 4 rotates about a rotation axis that is substantially perpendicular to the holding surface 4a, using power transmitted from the rotary drive source 12 via the spindle 10. Furthermore, the rotation axis of the rotary table 4 is located at the center of the rotary table 4.

[0042] A connecting part 14 is installed around the center of the main shaft 10. The connecting part 14 is connected to the suction source 18, such as an injector, via a flow path 16 formed by a hose, pipe, etc. The holding surface 4a of the rotary table 4 (the suction surface 8a of the holding member 8) is connected to the suction source 18 via the hole of the holding member 8, a flow path (not shown) formed inside the frame 6 and the main shaft 10, the connecting part 14, and the flow path 16.

[0043] A camera (imaging unit) 20 for photographing the wafer 11 is disposed above the rotary table 4. For example, a visible light camera having an imaging element that receives visible light and converts it into an electrical signal, or an infrared camera having an imaging element that receives infrared light and converts it into an electrical signal, can be used as the camera 20. The camera 20 photographs an object positioned in the imaging area 22 of the camera 20 to generate an image of the photographed object.

[0044] Each component of the notch detection device 2 (rotation drive source 12, suction source 18, camera 20, etc.) is connected to the control unit (control unit, control device) 24. The control unit 24 outputs control signals to the components of the notch detection device 2, thereby controlling the operation of the notch detection device 2.

[0045] For example, the control unit 24 is configured as a computer, including an arithmetic unit (not shown) for performing calculations required to operate the notch detection device 2, and a storage unit 26 for storing various information (data, programs, etc.). The arithmetic unit is configured to include a processor such as a CPU (Central Processing Unit). In addition, the storage unit 26 is configured to include various memories that constitute main storage devices, auxiliary storage devices, etc.

[0046] For example, the notch detection device 2 can be built into the various processing or inspection devices described above. Alternatively, the notch detection device 2 can be installed independently of the processing or inspection device and connected to it. Furthermore, the orientation of the wafer 11 processed by the processing device and the orientation of the wafer 11 inspected by the inspection device can be adjusted according to the position of the notch 11c detected by the notch detection device 2.

[0047] Figure 3 This is a flowchart illustrating the operation of the notch detection device 2 when detecting the notch 11c of the wafer 11. Hereinafter, refer to... Figure 2 , Figure 3 Specific examples of the notch detection method of this embodiment will be described.

[0048] First, a placement process (step S1) is performed, in which the wafer 11 is placed on the rotary table 4. The wafer 11 is transported to the rotary table 4 by a conveying mechanism (not shown), etc. Figure 2 As shown, for example, wafer 11 is arranged on rotary table 4 with its front side 11a exposed upwards and its back side 11b facing the holding surface 4a. When the attraction force (negative pressure) of attraction source 18 is applied to attraction surface 8a in this state, wafer 11 is attracted and held by rotary table 4.

[0049] Furthermore, the diameter of the holding surface 4a of the rotary stage 4 is smaller than the diameter of the wafer 11. Additionally, the wafer 11 is arranged such that its center is approximately aligned with the center of the rotary stage 4. Therefore, when the wafer 11 is held by the rotary stage 4, the outer periphery of the wafer 11 is positioned radially outward from the outer periphery of the holding surface 4a. In other words, the wafer 11 is arranged such that its outer periphery protrudes from the holding surface 4a.

[0050] Furthermore, the position of the camera 20 is adjusted such that the shooting area 22 of the camera 20 is located radially outward from the holding surface 4a of the rotary stage 4. Also, when the wafer 11 is held by the rotary stage 4, a portion of the outer periphery of the wafer 11 is positioned within the shooting area 22 of the camera 20.

[0051] However, the diameter of the holding surface 4a of the rotary stage 4 can be the same as or larger than the diameter of the wafer 11. In these cases, the holding surface 4a of the rotary stage 4 can be used to support the entire wafer 11. For example, when holding a thin wafer 11 that is prone to bending, it is preferable to use the holding surface 4a of the rotary stage 4 to support the entire wafer 11.

[0052] Next, the imaging process (step S2) is performed, in which the rotary table 4 is rotated while the camera 20 captures an image of the outer periphery of the wafer 11. Specifically, the control unit 24 outputs a control signal to the rotation drive source 12, causing the rotary table 4 to rotate at a predetermined speed. Additionally, the control unit 24 outputs a control signal to the camera 20, which then captures multiple images of the outer periphery of the wafer 11 positioned within the imaging area 22 at predetermined time intervals.

[0053] As the rotary stage 4 rotates, different regions of the outer periphery of the wafer 11 are sequentially positioned within the imaging area 22 of the camera 20. Furthermore, the outer periphery of the wafer 11 is continuously photographed along its circumference. This results in the acquisition of multiple images showing magnified versions of different regions of the outer periphery of the wafer 11.

[0054] Furthermore, the camera 20 continues to photograph the outer periphery of the wafer 11 until the wafer 11 has rotated one full revolution by the rotation of the rotary table 4. The shooting interval of the camera 20 is set according to the rotation speed of the rotary table 4 so that the entire outer periphery of the wafer 11 can be photographed by the camera 20. Therefore, when the imaging process is performed, one or more images including the notch 11c of the wafer 11 are obtained.

[0055] An example of an image of chip 11 acquired during the imaging process is shown below. Figure 4 (A) Figure 4 (E). Figure 4 (A) Figure 4 (E) shows images 30A to 30E obtained by repeatedly photographing the outer periphery of the chip 11 using the camera 20. Images 30A to 30E each include: a presence area (bright area) 32 corresponding to the area where the chip 11 is present; and a non-present area (dark area) 34 corresponding to the area where the chip 11 is not present. However, sometimes the brightness and darkness may be reversed depending on the shooting conditions of the chip 11 (lighting conditions, camera 20 sensitivity, etc.), with the area where the chip 11 is present becoming a dark area and the area where the chip 11 is not present becoming a bright area.

[0056] Figure 4 Image (A) is an image diagram showing the area where the notch 11c is not located (image 30A, first image). When the notch 11c is not positioned in the shooting area 22 of the camera 20 (see reference...) Figure 2 When the wafer 11 is photographed in a state where it is not visible, an image 30A is obtained that does not contain the pattern corresponding to the notch 11c. In the image 30A, the arc-shaped outline 36 corresponding to the outer periphery of the wafer 11 is represented by the boundary line between the present region 32 and the non-present region 34.

[0057] Figure 4 Image (B) is an image diagram showing the area where the notch 11c exists (image 30B, second image). When the notch 11c is positioned in the shooting area 22 of the camera 20 (see reference). Figure 2 When the wafer 11 is photographed in a certain state, an image 30B is obtained that includes an irregularly shaped region 38A corresponding to the notch 11c. In the image 30B, in addition to the arc-shaped outline 36, the outline 40A of the irregularly shaped region 38A corresponding to the edge of the notch 11c is also shown.

[0058] Additionally, sometimes there are notches other than the notch 11c on the outer periphery of the wafer 11. For example, when the wafer 11 is cut along the spacer 13 (see reference) using a cutting tool mounted on a cutting device. Figure 1 During the cutting process, the outer periphery of the wafer 11 sometimes fragments and scatters, resulting in defects on the outer periphery of the wafer 11. In addition, when the wafer 11 is ground using a grinding wheel mounted on a grinding device, defects sometimes occur on the thinned outer periphery of the wafer 11 due to the pressure applied to the wafer 11 by the grinding wheel.

[0059] As described above, sometimes, due to various reasons, cuts other than the notch 11c are accidentally formed on the outer periphery of the wafer 11. In these cases, when the camera 20 continuously photographs the outer periphery of the wafer 11, images containing the cuts other than the notch 11c are obtained. Hereinafter, as an example, cases where not only the notch 11c is formed on the outer periphery of the wafer 11, but also semi-circular defects, fan-shaped defects, and rectangular defects will be described.

[0060] Figure 4 Image (C) is an image 30C (third image) showing a region with a semi-circular defect. When a semi-circular defect exists on the outer periphery of the wafer 11, image 30C, including an irregularly shaped region 38B corresponding to the semi-circular defect, is acquired by camera 20. Furthermore, in image 30C, in addition to the arc-shaped outline 36, the outline 40B of the irregularly shaped region 38B corresponding to the edge of the semi-circular defect is also shown.

[0061] Figure 4 Image (D) is an image 30D (fourth image) showing a region with a fan-shaped defect. When a fan-shaped defect exists on the outer periphery of the wafer 11, image 30D is acquired by camera 20, showing an irregularly shaped region 38C corresponding to the fan-shaped defect. Furthermore, in image 30D, in addition to the arc-shaped outline 36, the outline 40C of the irregularly shaped region 38C corresponding to the edge of the fan-shaped defect is also shown.

[0062] Figure 4 Image (E) is an image 30E (5th image) showing a region with a rectangular defect. When a rectangular defect exists on the outer periphery of the wafer 11, image 30E, including an irregularly shaped region 38D corresponding to the rectangular defect, is acquired by camera 20. Furthermore, in image 30E, in addition to the arc-shaped outline 36, the outline 40D of the irregularly shaped region 38D corresponding to the edge of the rectangular defect is also shown.

[0063] As described above, the outer periphery of the wafer 11 is photographed multiple times using the camera 20, thereby generating an image of the outer periphery of the wafer 11 (images 30A-30E). Furthermore, the images acquired by the camera 20 are input to the control unit 24 (see reference). Figure 2 ), stored in storage unit 26 (memory).

[0064] Additionally, when placing the wafer 11 on the rotary table 4 (see reference) Figure 2 When the notch 11c on the wafer is positioned within a specified area, the camera 20 can capture only that area. In this case, it is not necessary to capture the entire outer periphery of the wafer 11, thus reducing the workload and time required for the capture process.

[0065] Next, a contour data acquisition process (step S3) is performed, in which contour data containing the coordinates of the contour of the wafer 11 is acquired based on images 30A to 30E.

[0066] In the contour data acquisition process, the control unit 24 (refer to...) Figure 2 Image processing, such as edge detection, is performed on images 30A to 30E respectively. As a result, contours 36 and 40A to 40D contained in images 30A to 30E are extracted, and the coordinates of contours 36 and 40A to 40D are determined. Furthermore, the coordinates of contours 36 and 40A to 40D correspond to the coordinates of the contour (outer periphery) of wafer 11. The contour data containing the coordinates of the contour of wafer 11 is then input to the control unit 24 and stored in the storage unit 26.

[0067] Next, a hypothetical circle calculation process (step S4) is performed, in which a hypothetical circle approximating the contour of wafer 11 is calculated based on the coordinates of the contour of wafer 11. In this hypothetical circle calculation process, contour 36 is approximated by a circle based on the coordinates of contour 36 extracted in the contour data acquisition process. Furthermore, there are no restrictions on the approximation method; for example, an approximation formula approximating contour 36 can be derived using the least squares method. The circle represented by this approximation formula corresponds to the hypothetical circle of the contour of wafer 11.

[0068] Figure 5 (A) is an image diagram showing an image 30A displaying an imaginary circle 42. Figure 5 Image (B) is an image 30B showing an imaginary circle 42. When images 30A and 30B are superimposed on the imaginary circle 42, the imaginary circle 42 is displayed along the outline 36. Additionally, in Figure 5 (A) and Figure 5 In (B), the imaginary circle 42 is shown by a dashed line.

[0069] Images 30A to 30E (reference) were examined respectively. Figure 4 (A) Figure 4The imaginary circle 42 is calculated (E). Furthermore, the imaginary circle (approximate form) calculated in the imaginary circle calculation process is stored in the control unit 24 (see reference). Figure 2 Storage section 26.

[0070] Next, an irregular shape region determination process (step S5) is performed to determine whether an irregular shape region exists on the outer periphery of the wafer 11. In addition, an irregular shape region is an irregular region on the outer periphery of the wafer 11 in which the outline of the wafer 11 is offset from the imaginary circle 42 by a certain degree or more.

[0071] In the irregular shape region determination process, the coordinates of the outline of the wafer 11 stored in the storage unit 26 of the control unit 24 (coordinates of outline 36, 40A~40D, reference) are read out. Figure 4 (A) Figure 4 The control unit 24 calculates the radial distance between the outline of the wafer 11 and the imaginary circle 42. This distance corresponds to the offset between the imaginary circle 42, which approximates the outline of the wafer 11, and the actual outline of the wafer 11.

[0072] The offset of the contour is calculated sequentially along contours 36, 40A to 40D. Furthermore, the control unit 24 determines whether there is an irregularly shaped region on the outer periphery of the wafer 11 based on the calculated offset.

[0073] For example, in the storage section 26 of the control section 24 (see reference) Figure 2 The control unit 24 stores a preset offset reference value and compares the calculated offset with the reference value sequentially. Furthermore, if the offset is above the reference value (or greater than the reference value), the control unit 24 determines that an irregularly shaped region exists, and if the offset is less than the reference value (or less than the reference value), the control unit 24 determines that an irregularly shaped region does not exist.

[0074] Specifically, in the area on the outer periphery of the wafer 11 where no notch 11c or defect is formed (see reference) Figure 4 In (A), the outline of wafer 11 is approximately aligned with the position of imaginary circle 42, with a small offset. As a result, it is determined that there are no irregularly shaped regions in this area. On the other hand, in the area where a notch 11c or a defect is formed on the outer periphery of wafer 11 (see reference...) Figure 4 (B) Figure 4 In (E), the outlines 40A to 40D of the wafer 11 do not follow the imaginary circle 42 and have a large offset. As a result, it is determined that there is an irregularly shaped region in this area.

[0075] Furthermore, there are no restrictions on the specific method for determining whether there are irregularly shaped regions. For example, the reference value for comparison with the offset of the wafer 11's contour can also be automatically set. Specifically, after calculating the offset of the contour sequentially along contours 36, 40A to 40D, the reference value can be set as an integer multiple (more than 2 times) of the center value of the calculated offset. Thus, a reference value that takes into account the deviation of the wafer 11's contour shape can be set for each wafer 11.

[0076] Next, the first notch determination process (step S6) is performed, and the non-circular region is determined to be a notch 11c based on the radial distance between the imaginary circle 42 and the front end of the non-circular region on the wafer 11.

[0077] In the first notch determination process, the control unit 24 (see reference) Figure 2 Calculate the radial distance D between the imaginary circle 42 and the leading edge of the irregularly shaped region on the wafer 11. Specifically, the region on the outer periphery of the wafer 11 that is determined to be an irregularly shaped region in the irregularly shaped region determination process (irregularly shaped regions 38A to 38D, refer to...) Figure 4 (B) Figure 4 The distance D is calculated in (E) respectively. In addition, the front end of the irregular shape region corresponds to the location in the edge (outline 40A~40D) of the irregular shape region that is furthest from the imaginary circle 42 in the radial direction of the wafer 11.

[0078] Figure 6 (A) is an image diagram showing the distance D (distance D1) in the irregularly shaped region 38A of image 30B. Figure 6 Image (B) is an image showing the distance D (distance D2) in the irregularly shaped region 38B of image 30C. The leading edge of the irregularly shaped region 38A corresponds to the location closest to the center of the wafer 11 in the contour 40A. Similarly, the leading edge of the irregularly shaped region 38B corresponds to the location closest to the center of the wafer 11 in the contour 40B.

[0079] Furthermore, the control unit 24 determines whether the irregularly shaped region is a notch 11c based on the calculated distance D. Specifically, in the storage unit 26 of the control unit 24 (see...) Figure 2 The pre-stored contents of notch 11c (refer to) Figure 1 The depth of the notch 11c corresponds to the reference value of the distance D. In addition, the depth of the notch 11c is equivalent to the distance between the outer end of the notch 11c in the radial direction and the inner end of the notch 11 in the radial direction.

[0080] For example, the lower limit of the depth of the notch 11c that can actually be formed on the wafer 11 is stored in the storage unit 26 as a reference value. Furthermore, the control unit 24 compares the calculated distance D with the reference value, thereby determining whether the irregularly shaped region is a notch 11c.

[0081] Specifically, Figure 6 The irregularly shaped region 38A shown in (A) corresponds to the notch 11c, and its distance from D1 is above the lower limit of the depth of the notch 11c. Therefore, the control unit 24 determines that the irregularly shaped region 38A is the notch 11c. On the other hand, Figure 6 The irregularly shaped region 38B shown in (B) corresponds to a defect smaller than the notch 11c, and the distance D2 is less than the lower limit of the depth of the notch 11c. Therefore, the control unit 24 determines that the irregularly shaped region 38B is not the notch 11c.

[0082] Furthermore, the specific method for determining the notch 11c is not limited to the method described above. For example, the range (upper and lower limits) of the depth of the notch 11c that can actually be formed on the wafer 11 can be stored in the storage unit 26 as a reference value. In addition, the reference value of the depth of the notch 11c can also be automatically set in the same way as the irregular shape region determination process described above. In this case, the operation of presetting the reference value and storing it in the storage unit 26 is not required.

[0083] As described above, in the first notch determination process, the size of the irregularly shaped region formed on the outer periphery of the wafer 11 is used to determine whether the irregularly shaped region is a notch 11c. Therefore, it is possible to prevent defects with different sizes from the notch 11c from being mistakenly identified as notches 11c, and to correctly detect the notch 11c.

[0084] However, when the size of the defect formed on the outer periphery of the wafer 11 is close to the size of the notch 11c, sometimes even performing the first notch determination process cannot identify the notch 11c. For example, Figure 4 (D) and Figure 4 As shown in (E), if there are irregularly shaped regions 38C and 38D with the same size as the notch 11c on the outer periphery of the wafer 11, the irregularly shaped regions 38C and 38D will also be identified as the notch 11c, and the notch 11c cannot be identified.

[0085] Therefore, when there are multiple irregularly shaped regions that are determined to be notches 11c by the first notch determination process, other determination methods are used to further determine whether the irregularly shaped region is a notch 11c. Specifically, the following steps are performed: a sharpness calculation step (step S7) to calculate the sharpness of the irregularly shaped region; and a second notch determination step (step S8) to determine whether the irregularly shaped region is a notch 11c based on the sharpness of the irregularly shaped region.

[0086] Specifically, the control unit 24 first performs image processing on the images (images 30B, 30D, 30E) of the irregularly shaped regions that have been determined to be notches 11c by the first notch determination process, and quantifies the sharpness of each irregularly shaped region. For example, the control unit 24 performs corner detection processing on the images of the irregularly shaped regions using the Moravec corner detection method, and calculates the value (corner likelihood) corresponding to the sharpness of the front end of the irregularly shaped region.

[0087] Next, the control unit 24 determines whether the irregularly shaped region is a notch 11c based on the sharpness of the irregularly shaped region. Specifically, in the storage unit 26 of the control unit 24 (see reference...) Figure 2 The reference value for the angular likelihood is stored in advance. For example, the upper limit of the angular likelihood of a notch 11c that can actually be formed on the wafer 11 is stored in the storage unit 26 as the reference value. Furthermore, the control unit 24 compares the calculated angular likelihood with the reference value, thereby determining whether the irregularly shaped region is a notch 11c.

[0088] Figure 7 (A) is an image diagram showing the front end 44A of the irregularly shaped region 38A in image 30B. Figure 7 Image (B) is an image showing the front end 44B of the irregularly shaped region 38C in image 30D. (See image 30D for details.) Figure 7 (A) and Figure 7 As shown in (B), the front end 44A of the irregularly shaped region 38A does not contain an angle, while the front end 44B of the irregularly shaped region 38C does contain an angle. Therefore, the angle likelihoods are different in the irregularly shaped regions 38A and 38C.

[0089] Figure 7 The irregularly shaped region 38A shown in (A) corresponds to the notch 11c, and the angular likelihood of the irregularly shaped region 38A is below the upper limit of the angular likelihood of the notch 11c. Therefore, the control unit 24 determines that the irregularly shaped region 38A is the notch 11c. On the other hand, Figure 7 The irregularly shaped region 38C shown in (B) corresponds to a defect that is sharper than the notch 11c, and the angular likelihood of the irregularly shaped region 38C is greater than the upper limit of the angular likelihood of the notch 11c. Therefore, the control unit 24 determines that the irregularly shaped region 38C is not the notch 11c.

[0090] Furthermore, there are no restrictions on the specific method for determining the notch 11c. For example, the range (upper and lower limits) of the angular likelihood of the notch 11c that can actually be formed on the wafer 11 can be stored in the storage unit 26 as a reference value. In addition, the reference value of the angular likelihood can also be set automatically. For example, after calculating the angular likelihood of each region included in the irregular shape region in sequence, an integer multiple (more than 2 times) of the center value of the calculated angular likelihood can be set as the reference value of the angular likelihood.

[0091] As described above, in the second notch determination process, the sharpness of the irregularly shaped region formed on the outer periphery of the wafer 11 is used to determine whether the irregularly shaped region is a notch 11c. Therefore, it is possible to prevent defects with a sharpness different from that of the notch 11c from being mistakenly identified as notches 11c, and to correctly detect notches 11c.

[0092] However, if the leading edge of the defect formed on the outer periphery of the wafer 11 does not include a corner, sometimes even performing the second notch determination process cannot determine the notch 11c. For example, Figure 4 As shown in (E), the front end of the irregularly shaped region 38D is flat and does not contain any sharp corners. Therefore, it is easy to determine that the irregularly shaped region 38D is also the notch 11c, and it is difficult to determine the notch 11c.

[0093] Therefore, in cases where multiple irregularly shaped regions are identified as notches 11c in the second notch determination step, other determination methods are used to further determine whether the irregularly shaped region is a notch 11c. Specifically, the following steps are implemented: a similarity calculation step (step S9), which calculates the similarity between the first image obtained by segmenting the image containing the irregularly shaped region and the second image; a third notch determination step (step S10), which determines whether the irregularly shaped region is a notch 11c based on the similarity between the flipped images of the first image and the second image; and a fourth notch determination step (step S11), which determines whether the irregularly shaped region is a notch 11c based on the similarity between the first image and the second image.

[0094] In the similarity calculation process, the image containing the irregularly shaped region is first segmented into a first image and a second image by drawing a straight line along the radial direction of the wafer 11 to divide the irregularly shaped region into two parts (segmentation process). The following is a detailed description of the case where images 30B and 30E are segmented into the first image and the second image, respectively.

[0095] Figure 8Image (A) is an image diagram showing image 30B divided into a first image 48A and a second image 50A by a dividing line 46A. Image 30B is divided into the first image 48A and the second image 50A by a straight line, namely the dividing line 46A, that bisects the irregularly shaped region 38A along the radial direction of the wafer 11. As a result, the first image 48A shows a region 48Aa containing an irregularly shaped region 38A1 that corresponds to a portion of the divided irregularly shaped region 38A. Additionally, the second image 50A shows a region 50Aa containing an irregularly shaped region 38A2 that corresponds to a portion of the divided irregularly shaped region 38A.

[0096] Figure 8 Image (B) is an image diagram showing image 30E divided into a first image 48B and a second image 50B by dividing line 46B. Image 30E is divided into the first image 48B and the second image 50B by dividing line 46B, which bisects the irregularly shaped region 38D along the radial direction of wafer 11. As a result, the first image 48B shows region 48Ba, which contains irregularly shaped region 38D1, corresponding to a portion of the divided irregularly shaped region 38D. Additionally, the second image 50B shows region 50Ba, which contains irregularly shaped region 38D2, ​​corresponding to a portion of the divided irregularly shaped region 38D.

[0097] Furthermore, there are no limitations on the method for determining the straight line (dividing line) that divides the irregularly shaped region into two parts. For example, an imaginary circle 42 (refer to...) can be drawn along the radial direction of wafer 11. Figure 5 One of the multiple straight lines (candidate lines) connecting (B) and the edge of the irregularly shaped region is selected as the dividing line.

[0098] Specifically, the candidate line with the longest length or the candidate line with the largest moving average length among multiple candidate lines is set as the dividing line. In addition, the candidate line that passes through the midpoint of the two ends of the irregularly shaped region (the connection point between one end of the irregularly shaped region and the imaginary circle 42 and the connection point between the other end of the irregularly shaped region and the imaginary circle 42) can be set as the dividing line.

[0099] Furthermore, the dividing line can also be set based on the position and length of the candidate lines. Specifically, firstly, multiple points within the region adjacent to the irregularly shaped area in the imaginary circle 42 are set as reference points. Next, the product xL representing the position of each reference point is calculated by multiplying the position coordinate x by the length L of the candidate line passing through that reference point. Then, the sum of the values ​​xL of all reference points is divided by the sum of the lengths L of the multiple candidate lines. Thus, the position coordinate x′ of the reference point, weighted by the length of the candidate lines, is calculated. Finally, the candidate line passing through the reference point determined by the position coordinate x′ is set as the dividing line.

[0100] Alternatively, the position of the reference point can be represented by an angle θ with an imaginary circle 42 as the reference instead of the position coordinate x. In this case, the angle θ′ of the reference point is calculated using the length of the candidate line. Furthermore, the candidate line of the reference point determined by the angle θ′ is set as the dividing line.

[0101] Next, the second image 50A of image 30B and the second image 50B of image 30E are flipped. Specifically, the control unit 24 performs image processing on image 30B, flipping the second image 50A around the dividing line 46A. Similarly, the control unit 24 performs image processing on image 30E, flipping the second image 50B around the dividing line 46B. Thus, flipped images of the second image 50A and the second image 50B are generated.

[0102] Figure 9 (A) is an image diagram showing image 30B, which contains a flipped image 50A′ of the first image 48A and the second image 50A. Figure 9 (B) is an image diagram showing image 30E, which contains a flipped image 50B′ of the first image 48B and the second image 50B. Region 50Aa′ is shown in the flipped image 50A′, which contains a region 38A2 of a different shape than the region before flipping (see reference). Figure 8 The irregularly shaped region 38A2′ corresponds to (A). Additionally, region 50Ba′ is shown in the flipped image 50B′, which contains the irregularly shaped region 38D2 (see reference 38A2′) before flipping. Figure 8 The irregularly shaped region 38D2′ corresponding to (B)).

[0103] Next, the control unit 24 calculates the first similarity (first similarity calculation step) corresponding to the similarity between the flipped image of the first image and the second image. Specifically, the control unit 24 first compares the first image 48A and the flipped image 50A′ of image 30B through image processing and quantifies the similarity between the two.

[0104] Furthermore, there are no restrictions on the image processing used to compare the first image 48A and the flipped image 50A′. For example, the similarity between region 48Aa contained in the first image 48A and region 50Aa′ contained in the flipped image 50A′ can be numerically calculated using similarity calculation methods employed in template matching, etc. The specific calculation method for similarity can be freely chosen, such as using SSD (Sum of Squared Difference), SAD (Sum of Absolute Difference), NCC (Normalized Cross Correlation), etc.

[0105] Similarly, the control unit 24 compares the first image 48B and the flipped image 50B′ of image 30E through image processing, and quantifies the similarity between the two. As a result, it obtains a numerical value representing the similarity between the first image 48A contained in image 30B and the flipped image 50A′, and a numerical value representing the similarity between the first image 48B contained in image 30E and the flipped image 50B′. This numerical value represents the first similarity.

[0106] Next, the control unit 24 calculates a second similarity corresponding to the similarity between the first image and the second image (second similarity calculation process). Specifically, the control unit 24 performs image processing on the first image 48A and the second image 50A of image 30B (refer to...). Figure 8 The similarity between the two images is numerically calculated by comparing the first image 48B and the second image 50B of image 30E. Similarly, the control unit 24 performs image processing to compare the first image 48B and the second image 50B of image 30E (refer to image 48B). Figure 8 The similarity between the two is numericalized by comparing (B) with (B).

[0107] The result is that a numerical value representing the similarity between the first image 48A and the second image 50A contained in image 30B, and a numerical value representing the similarity between the first image 48B and the second image 50B contained in image 30E, are obtained. The numerical value representing this similarity is equivalent to the second similarity.

[0108] Furthermore, there are no restrictions on the comparison method between the first image and the second image; for example, the similarity can be calculated using the same method as the first similarity calculation step. Specifically, in image 30B, the similarity between region 48Aa contained in the first image 48A and region 50Aa contained in the second image 50A is numerically calculated. Additionally, in image 30D, the similarity between region 48Ba contained in the first image 48B and region 50Ba contained in the second image 50B is numerically calculated.

[0109] Next, the control unit 24 determines whether the irregularly shaped region is a notch 11c based on the first similarity score (third notch determination process). For example, a preset reference value for the first similarity score is stored in the storage unit 26 of the control unit 24, and the control unit 24 compares the calculated first similarity score with the reference value. Furthermore, the control unit 24 determines whether the irregularly shaped region is a notch 11c based on the relationship between the first similarity score and the reference value. Specifically, if the similarity between the first image and the flipped image is greater than or equal to the reference value, the control unit 24 determines that the irregularly shaped region is a notch 11c.

[0110] like Figure 9As shown in (A), the first image 48A contains region 48Aa showing an arc-shaped pattern (irregular shape region 38A1), and the flipped image 50A′ contains region 50Aa′ showing an arc-shaped pattern (irregular shape region 38A2′). Furthermore, the two patterns are approximately the same size and are similar. Therefore, the first image 48B and the flipped image 50B′ contain patterns similar to each other, with the first image 48B showing a higher degree of similarity to the flipped image 50B′.

[0111] In addition, such as Figure 9 As shown in (B), the first image 48B contains region 48Ba showing a rectangular pattern (irregular shape region 38D1), and the flipped image 50B' contains region 50Ba' showing a rectangular pattern (irregular shape region 38D2'). Furthermore, the width and height of the two patterns are approximately the same, indicating they are similar. Therefore, the first image 48B and the flipped image 50B' contain patterns that are similar to each other, and the first image 48B and the flipped image 50B' have a high degree of similarity.

[0112] Therefore, the similarity between the first image 48A and the flipped image 50A′, and the similarity between the first image 48B and the flipped image 50B′, are both greater than the baseline value. As a result, the irregularly shaped region 38A contained in image 30B and the different irregularly shaped regions 38D contained in image 30E are both identified as notches 11c.

[0113] Furthermore, the control unit 24 determines whether the irregularly shaped region is a notch 11c based on the second similarity (fourth notch determination process). For example, a preset reference value for the second similarity is stored in the storage unit 26 of the control unit 24, and the control unit 24 compares the calculated second similarity with the reference value. The control unit 24 then determines whether the irregularly shaped region is a notch 11c based on the relationship between the second similarity and the reference value. Specifically, if the similarity between the first image and the second image is below (or less than) the reference value, the control unit 24 determines that the irregularly shaped region is a notch 11c.

[0114] like Figure 8 As shown in (A), the first image 48A contains region 48Aa showing an arc-shaped pattern (irregular shape region 38A1), and the second image 50A contains region 50Aa showing an arc-shaped pattern (irregular shape region 38A2). However, the two patterns differ in angle by approximately 90° and are not similar. Therefore, neither the first image 48A nor the second image 50A contains patterns similar to each other, and the similarity between the first image 48A and the second image 50A is low.

[0115] On the other hand, such as Figure 8As shown in (B), the first image 48B contains region 48Ba showing a rectangular pattern (irregular shape region 38D1), and the second image 50B contains region 50Ba showing a rectangular pattern (irregular shape region 38D2). Furthermore, the width and height of the two patterns are approximately the same, indicating they are similar. Therefore, the first image 48B and the second image 50B contain patterns that are similar to each other, and the first image 48B and the second image 50B have a high degree of similarity.

[0116] Therefore, the similarity between the first image 48A and the second image 50A is lower than the benchmark value. As a result, the irregularly shaped region 38A contained in image 30B is determined to be the notch 11c. On the other hand, the similarity between the first image 48B and the second image 50B is higher than the benchmark value. As a result, the irregularly shaped region 38D contained in image 30E is determined not to be the notch 11c.

[0117] As described above, the third notch determination process eliminates asymmetrical irregularly shaped regions from the candidates for notch 11c, and the fourth notch determination process eliminates rectangular irregularly shaped regions from the candidates for notch 11c. For example, in this embodiment, the irregularly shaped region 38D contained in image 30E is symmetrical, so although it was not eliminated from the candidates for notch 11c in the third notch determination process, it was eliminated from the candidates for notch 11c in the subsequent fourth notch determination process. As a result, the irregularly shaped region 38A contained in image 30B (see reference) is determined. Figure 4 (B) represents notch 11c, and the irregularly shaped regions 38B to 38D contained in images 30C to 30E (refer to...) Figure 4 (C)~ Figure 4 (E) indicates the cut portion outside the notch 11c.

[0118] As described above, in the notch detection method of this embodiment, the size or shape of the irregularly shaped region formed on the outer periphery of the wafer 11 is used to determine whether the irregularly shaped region is a notch 11c. Therefore, even if a notch other than a notch 11c is formed on the outer periphery of the wafer 11, it is not easy to mistakenly identify the notch as a notch 11c, and the notch 11c can be detected correctly.

[0119] The above-mentioned notch detection method is implemented by the control unit 24 (refer to...). Figure 2 It is implemented automatically. Specifically, the storage unit 26 of the control unit 24 stores a series of actions described for detecting the notch 11c (see reference). Figure 3 The procedure in steps S2 to S11. Furthermore, when the wafer 11 is placed on the rotary table 4 (step S1), the control unit 24 reads the program from the storage unit 26 and executes it to detect the notch 11c. Thus, the detection of the notch 11c is performed easily and quickly.

[0120] Furthermore, in this embodiment, an example of performing four notch inspection processes (notch inspection processes 1 to 4) sequentially has been described, but the number of notch inspection processes can be appropriately changed. Specifically, it is also possible to perform only one, two, or three of the notch inspection processes 1 to 4. In addition, when multiple notch inspection processes are performed, the order of the notch inspection processes can be freely set.

[0121] In addition, the structure and method of the above-described embodiments can be appropriately modified and implemented as long as they do not depart from the scope of the purpose of the present invention.

Claims

1. A notch detection method for detecting notches formed on the outer periphery of a disk-shaped wafer and indicating the crystal orientation of the wafer, characterized in that, The notch detection method includes the following steps: The placement process involves placing the wafer on a rotary table and positioning a portion of the wafer's outer periphery within the camera's imaging area. The imaging process involves rotating the rotary worktable while using the camera to take multiple images of the wafer, thereby obtaining an image of the outer periphery of the wafer. The contour data acquisition process acquires contour data based on the image, and the contour data includes the coordinates of the wafer's contour. The imaginary circle calculation process calculates an imaginary circle approximating the outline of the wafer based on the coordinates of the wafer's outline. The irregular shape region determination process determines whether an irregular shape region exists on the outer periphery of the wafer based on the radial distance between the imaginary circle and the wafer's outline on the wafer. The first notch determination process determines whether the irregularly shaped area is a notch based on the radial distance between the imaginary circle and the front end of the irregularly shaped area on the wafer. The sharpness calculation process involves performing corner detection processing on the image to calculate the angular likelihood of the irregularly shaped region. as well as The second indentation determination process determines whether the irregularly shaped area is an indentation based on the likelihood of the angle.

2. A notch detection method for detecting notches formed on the outer periphery of a disk-shaped wafer and indicating the crystal orientation of the wafer, characterized in that, The notch detection method includes the following steps: The placement process involves placing the wafer on a rotary table and positioning a portion of the wafer's outer periphery within the camera's imaging area. The imaging process involves rotating the rotary worktable while using the camera to take multiple images of the wafer, thereby obtaining an image of the outer periphery of the wafer. The contour data acquisition process acquires contour data based on the image, and the contour data includes the coordinates of the wafer's contour. The imaginary circle calculation process calculates an imaginary circle approximating the outline of the wafer based on the coordinates of the wafer's outline. The irregular shape region determination process determines whether an irregular shape region exists on the outer periphery of the wafer based on the radial distance between the imaginary circle and the wafer's outline on the wafer. The first notch determination process determines whether the irregularly shaped area is a notch based on the radial distance between the imaginary circle and the front end of the irregularly shaped area on the wafer. The segmentation process divides the image into a first image and a second image by using a straight line that divides the irregularly shaped region into two parts along the radial direction of the wafer. The first similarity calculation step calculates a first similarity corresponding to the similarity between the first image and the flipped image, wherein the flipped image is obtained by flipping the second image around the straight line as the axis. The second similarity calculation step calculates the second similarity corresponding to the similarity between the first image and the second image; The third notch determination process determines whether the irregularly shaped area is a notch based on the first similarity. as well as The fourth notch determination process determines whether the irregularly shaped area is a notch based on the second similarity.

3. A notch detection method for detecting notches formed on the outer periphery of a disk-shaped wafer and indicating the crystal orientation of the wafer, characterized in that, The notch detection method includes the following steps: The placement process involves placing the wafer on a rotary table and positioning a portion of the wafer's outer periphery within the camera's imaging area. The imaging process involves rotating the rotary worktable while using the camera to take multiple images of the wafer, thereby obtaining an image of the outer periphery of the wafer. The contour data acquisition process acquires contour data based on the image, and the contour data includes the coordinates of the wafer's contour. The imaginary circle calculation process calculates an imaginary circle approximating the outline of the wafer based on the coordinates of the wafer's outline. The irregular shape region determination process determines whether an irregular shape region exists on the outer periphery of the wafer based on the radial distance between the imaginary circle and the wafer's outline on the wafer. The first notch determination process determines whether the irregularly shaped area is a notch based on the radial distance between the imaginary circle and the front end of the irregularly shaped area on the wafer. The sharpness calculation process calculates the sharpness of the irregularly shaped region. The second notch determination process determines whether the irregularly shaped area is a notch based on the sharpness of the area. The segmentation process divides the image into a first image and a second image by using a straight line that divides the irregularly shaped region into two parts along the radial direction of the wafer. The first similarity calculation step calculates a first similarity corresponding to the similarity between the first image and the flipped image, wherein the flipped image is obtained by flipping the second image around the straight line as the axis. The second similarity calculation step calculates the second similarity corresponding to the similarity between the first image and the second image; The third notch determination process determines whether the irregularly shaped area is a notch based on the first similarity. as well as The fourth notch determination process determines whether the irregularly shaped area is a notch based on the second similarity.

Images