A wafer defect detection method, device and medium

Abstract

This disclosure provides a wafer defect detection method, apparatus, and medium. The wafer defect detection method includes: acquiring an edge image of a wafer; determining a region to be identified containing defects based on image features of the edge image; identifying two measured boundaries in the radial direction of the wafer in the region to be identified based on image features of the region to be identified; determining two target boundaries in the radial direction of the wafer in the region to be identified based on the two measured boundaries; and determining defect information in the region to be identified based on the two target boundaries and the image features of the region to be identified. The wafer defect detection method, apparatus, and medium provided in this disclosure can avoid the problem of missed defect detection.

Description

Technical Field

[0001] This invention relates to the field of semiconductor processing technology, and in particular to a method, apparatus and medium for detecting wafer defects. Background Technology

[0002] During wafer manufacturing, defects can form on the wafer at any stage of the process. As semiconductor technology advances and semiconductor devices become increasingly miniaturized, wafer defects have a growing impact on the semiconductor process. Therefore, during wafer manufacturing, defect detection is necessary. The detection results are used to analyze the causes of defects and adjust the production process or equipment to reduce defect generation.

[0003] In the semiconductor industry, one method for detecting defects in wafer edge regions is to acquire an image of the wafer edge, crop out the area to be identified that includes the defect, and determine the boundary of the edge defect by comparing the grayscale of the image within the area to be identified. Based on the characteristic parameters of different defects, the type of defect is identified and the defect is graded.

[0004] However, the above-mentioned defect detection methods have the following problems: In the defect judgment process, the accuracy of identifying the wafer edge in the image will directly affect the defect judgment. For example, if the defect is too large, the edge of the silicon wafer cannot be accurately captured, resulting in the problem of missed defect detection. Summary of the Invention

[0005] This disclosure provides a wafer defect detection method, apparatus, and medium that can avoid the problem of missed defect detection.

[0006] The technical solutions provided in this disclosure are as follows:

[0007] In a first aspect, embodiments of this disclosure provide a wafer defect detection method, comprising:

[0008] Obtain the edge image of the wafer;

[0009] Based on the image features of the edge image, determine the region to be identified that has defects;

[0010] Based on the image features in the region to be identified, two measured boundaries of the wafer radial direction in the region to be identified are identified;

[0011] Based on the two measured boundaries, two target boundaries in the wafer radial direction within the region to be identified are determined;

[0012] Based on the image features of the two target boundaries and the region to be identified, the defect information within the region to be identified is determined.

[0013] For example, identifying two measured boundaries of the wafer radial direction in the region to be identified based on image features in the region to be identified specifically includes:

[0014] In the grayscale image corresponding to the region to be identified, wafer images and non-wafer images with grayscale difference values ​​greater than the grayscale threshold are identified;

[0015] Based on the morphological features of the wafer image and the non-wafer image, a reference region of a wafer is selected. The two boundaries of the reference region of the wafer have an extension length along the edge of the wafer in the region to be identified that is greater than or equal to a first percentage threshold.

[0016] The two boundaries of the reference region of the wafer are defined as the two measured boundaries.

[0017] For example, determining two target boundaries in the wafer radial direction within the region to be identified based on the two measured boundaries specifically includes:

[0018] If the measured distance between the two measured boundaries in the wafer radial direction is less than or equal to the maximum distance threshold and greater than or equal to the minimum distance threshold, the two measured boundaries are determined to be the two target boundaries.

[0019] For example, determining two target boundaries in the wafer radial direction within the region to be identified based on the two measured boundaries specifically includes:

[0020] If the measured distance between the two measured boundaries in the wafer radial direction is greater than the maximum distance threshold or less than the minimum distance threshold, verify whether one of the two measured boundaries is a target boundary.

[0021] If the verification result shows that one of the two measured boundaries is a target boundary, then based on the verified target boundary, obtain the other target boundary;

[0022] If the verification result is that neither of the two measured boundaries is the target boundary, a preset custom boundary is used as the target boundary, and the other target boundary is obtained based on the custom boundary.

[0023] For example, the two measured boundaries include a first measured boundary and a second measured boundary; verifying whether one of the two measured boundaries is a target boundary specifically includes:

[0024] The first measured boundary is offset by a first distance in a direction closer to the second measured boundary to obtain the first virtual boundary;

[0025] If the length proportion of the wafer image corresponding to the first virtual boundary in the region to be identified is greater than or equal to the second proportion threshold, the first measured boundary is determined to be a target boundary.

[0026] If the length proportion of the wafer image corresponding to the first virtual boundary in the region to be identified is less than the second proportion threshold, the second measured boundary is shifted to the direction closer to the first measured boundary by the first distance to obtain the second virtual boundary;

[0027] If the length proportion of the wafer image corresponding to the second virtual boundary in the region to be identified is greater than or equal to the second proportion threshold, the second measured boundary is determined to be a target boundary.

[0028] For example, obtaining another target boundary based on a verified target boundary specifically includes:

[0029] Using one of the verified target boundaries as a reference, a boundary is obtained by offsetting it by a second distance in the direction of another measured boundary on its opposite side, which is taken as the other target boundary.

[0030] For example, obtaining another target boundary based on the custom boundary specifically includes:

[0031] Using the custom boundary as a reference, a boundary is obtained by offsetting it by a second distance in the direction of another measured boundary on its opposite side, which is used as the other target boundary.

[0032] For example, when the measured distance of the wafer image in the region to be identified along the wafer radial direction is within a distance threshold, the second distance is the measured distance of the wafer image in the region to be identified along the wafer radial direction.

[0033] When the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is not within the distance threshold, the second distance is a preset distance value.

[0034] Secondly, embodiments of this disclosure provide a wafer defect detection apparatus, including a processor and a memory; the processor is used to execute instructions stored in the memory to implement the wafer defect detection method as described above.

[0035] Thirdly, embodiments of this disclosure provide a computer storage medium storing at least one instruction, which is executed by a processor to implement the wafer defect detection method described above.

[0036] The beneficial effects of the embodiments disclosed herein are as follows:

[0037] In the above scheme, after acquiring the edge image of the wafer, the region to be identified for defects is determined based on the image features of the edge image. Based on the image features of the region to be identified, two measured boundaries in the radial direction of the wafer are identified. Based on the two measured boundaries, two target boundaries are further determined. Based on the two target boundaries and the image features of the region to be identified, the feature parameters of the defects contained in the region to be identified are obtained. These parameters are compared with the feature parameters of different types of defects and / or different levels of defects to identify the type of the current defect in the region to be identified and to classify the current defect. This determines the defect information in the region to be identified, which can avoid the problem of missed defect detection due to failure of defect boundary identification. Attached Figure Description

[0038] Figure 1 A flowchart illustrating the wafer defect detection method in this embodiment of the present disclosure;

[0039] Figure 2 A schematic diagram illustrating step S033 of the wafer defect detection method in an embodiment of this disclosure;

[0040] Figure 3 One of the schematic diagrams illustrating step S042 of the wafer defect detection method in the embodiments of this disclosure;

[0041] Figure 4 This is the second schematic diagram illustrating step S042 of the wafer defect detection method in this embodiment of the present disclosure;

[0042] Figure 5 A schematic diagram illustrating step S044 of the wafer defect detection method in an embodiment of this disclosure;

[0043] Figures 6A to 6D This indicates several types of defect images that can be detected by the wafer defect detection method in this embodiment of the present disclosure. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0045] Unless otherwise defined, the technical or scientific terms used in this disclosure shall have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms “first,” “second,” and similar terms used in this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an,” “a,” or “the,” and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms “including,” “comprising,” or “containing,” and similar terms mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. The terms “connected,” “linked,” or similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms “upper,” “lower,” “left,” and “right,” etc., are used only to indicate relative positional relationships, and these relative positional relationships may change accordingly when the absolute position of the described objects changes.

[0046] The features such as "parallel," "perpendicular," and "identical" used in the embodiments of this disclosure include features in the strict sense of "parallel," "perpendicular," and "identical," as well as cases where "approximately parallel," "approximately perpendicular," and "approximately identical" include certain tolerances. Taking into account the measurement and the tolerances associated with the measurement of a specific quantity (e.g., limitations of the measurement system), they represent the acceptable deviation range for a specific value as determined by a person skilled in the art. For example, "approximately" can mean within one or more standard deviations, or within 3% or 5% of said value.

[0047] Furthermore, throughout this document, unless otherwise defined, the terms “substantially,” “essentially,” “approximately,” and “about” are used to describe and explain small variations. When used with an event or situation, these terms can cover situations where the event or situation occurs precisely or approximately. For example, when used with a numerical value, these terms can include a range of variation of the numerical value less than or equal to 10%, such as less than or equal to ±5%, less than or equal to ±4%, less than or equal to ±3%, less than or equal to ±2%, less than or equal to ±1%, less than or equal to ±0.5%, less than or equal to ±0.1%, or less than or equal to ±0.05%. The term “substantially coplanar” can refer to two surfaces arranged along the same plane within a micrometer range, for example, within 40 μm, 30 μm, 20 μm, 10 μm, or 1 μm.

[0048] After obtaining a single-crystal silicon ingot through the Czochralski method, the single-crystal silicon ingot can be processed by cutting, grinding, polishing, cleaning and other processes to obtain a polished wafer. In some cases, an epitaxial layer of single-crystal silicon can be formed on the surface of the polished wafer by chemical vapor deposition to obtain an epitaxial wafer.

[0049] Regardless of whether it is a polished wafer or an epitaxial wafer, edge defects have an increasingly significant impact on subsequent integrated circuit manufacturing processes and product yield. Based on medium- and long-term observations and statistics of the wafer production process, it has been found that approximately 90% of the edge defects are chip defects, crack defects, and scratch defects.

[0050] Specifically, for the three types of defects mentioned above, Chip defects are edge-bumping defects with variable angles, which can easily cause fragmentation risks in subsequent manufacturing processes, and this type of defect cannot be removed by rework; Crack defects are linear defects that extend along the longitudinal direction of the wafer, which are highly prone to fragmentation risks after heating, and this type is certain to be irremovable by rework; Scratch defects are mostly horizontally extending defects, and are mostly caused by malfunctions in edge polishing equipment, and most can be removed by rework.

[0051] Currently, silicon wafer defect detection technology involves scanning the wafer with a camera, selecting areas with significant grayscale value changes that include defects as the areas to be identified, determining the boundaries of edge defects by comparing the grayscale values ​​of the images within these areas, extracting feature parameters of different defects using software, identifying different types of defects using these feature parameters, and classifying the defects.

[0052] In the above defect determination process, the accuracy of identifying the wafer edge in the image directly affects the defect determination. For example, if the defect is too large, the edge of the silicon wafer cannot be accurately captured, leading to the problem of missed defect detection.

[0053] To address the aforementioned issues, this disclosure provides a wafer defect detection method, apparatus, and medium that can prevent missed defect detection.

[0054] like Figure 1 As shown, the wafer defect detection method provided in this embodiment includes:

[0055] Step S01: Obtain the edge image of the wafer;

[0056] Step S02: Based on the image features of the edge image, determine the region to be identified that has defects;

[0057] Step S03: Based on the image features in the region to be identified, identify two measured boundaries of the wafer radial direction in the region to be identified;

[0058] Step S04: Based on the two measured boundaries, determine the two target boundaries of the wafer radial direction in the region to be identified;

[0059] Step S05: Based on the image features of the two target boundaries and the region to be identified, determine the defect information within the region to be identified.

[0060] In the above scheme, after acquiring the edge image of the wafer, the defect to be identified region is determined based on the image features of the edge image. Based on the image features of the region to be identified, two measured boundaries in the radial direction of the wafer are identified. Based on the two measured boundaries, two target boundaries are further determined. Based on the two target boundaries and the image features of the region to be identified, the feature parameters of the defects contained in the region to be identified are obtained. These parameters are compared with the feature parameters of different types of defects and / or different levels of defects to identify the type of the current defect in the region to be identified and to classify the current defect. This method can determine the defect information in the region to be identified and avoid the defect omission problem caused by directly identifying the measured boundaries.

[0061] It should be noted that, in this application, the target boundary refers to the boundary line that is intended to reflect the actual boundary of the wafer.

[0062] In step S01 above, the wafer edge image can refer to the full circumference edge image of the wafer. A full circumference edge image of the wafer is an image showing the entire edge outline of the wafer. For example, a wafer typically has a notch, and a complete full circumference edge image can start from and end with a notch-shaped pattern. In other words, the wafer edge image is the portion of the image between two notch patterns in the complete image.

[0063] Edge images of a wafer can be acquired using specialized image acquisition equipment by placing the wafer on a table, ensuring that the wafer surface faces the lens of the image acquisition equipment, and ensuring that the acquired image fully covers the radially opposite edges of the entire wafer; rotating the wafer relative to the lens to scan the wafer surface to obtain the edge images of the wafer.

[0064] It should be noted that the above is merely an illustrative description of acquiring wafer edge images and is not intended to limit the scope of the invention. It is understood that those skilled in the art can acquire full-circumference edge images of wafers using other methods or systems, and the embodiments of the present invention will not be elaborated upon further.

[0065] In step S02 above, please refer to some examples. Figure 2As shown, the shape of the region A to be identified can be rectangular, and it can completely encompass the defects within it. For various edge defect types, due to their different formation mechanisms, their morphologies at the wafer edge also differ, and furthermore, the morphological features presented in the image are not entirely the same. Therefore, morphological features can directly characterize the shape, size, area, and other morphological parameters of defects at the wafer edge, providing a more accurate basis for defect type identification.

[0066] In some exemplary embodiments, step S03 specifically includes:

[0067] Step S031: In the grayscale image corresponding to the region to be identified, identify the wafer image A1 and the non-wafer image A2 where the grayscale difference is greater than the grayscale threshold;

[0068] Step S032: Based on the morphological features of the wafer image A1 and the non-wafer image A2, a reference region B of a wafer is selected. The two boundaries of the reference region B of the wafer have an extension length along the edge of the wafer in the region to be identified that is greater than or equal to a first percentage threshold.

[0069] Step S033: Determine the two boundaries of the reference region B of the wafer as the two measured boundaries C.

[0070] In step S031 above, the original image of the wafer edge can be binarized by setting the pixel grayscale values ​​in the image to the range of 0 to 255, thus obtaining the wafer edge image. Compared to the original image, the measured boundary C of the wafer can be distinguished more clearly. For example, in the non-wafer area, the grayscale is 0, while the grayscale in the wafer area is close to 150. Therefore, there will be a jump in grayscale value > 100 at the wafer boundary. Based on this, the wafer image and the non-wafer image in the area to be identified can be distinguished, so as to identify the edge contour of the wafer in the edge image.

[0071] In steps S032 and S033 above, such as Figure 2As shown, a first percentage threshold can be defined. A reference region B of a wafer can be selected from the region to be identified. The straight-line extension lengths of the two boundaries of this reference region B along the wafer's radial direction in the wafer edge extension direction are calculated as a percentage of the total length of the entire region to be identified along the wafer edge extension direction. If this percentage is greater than or equal to the first percentage threshold, the two boundaries of the selected reference region B of the wafer can be tentatively defined as the two measured boundaries C of the region to be identified. This ensures that edge defects of the wafer will not be identified as boundaries. For example, the first percentage threshold can be 10% to 30%, such as 10%, 20%, 30%, etc. However, it is not limited to this.

[0072] In some exemplary embodiments, step S04 specifically includes:

[0073] Step S041: If the measured distance between the two measured boundaries C in the wafer radial direction is less than or equal to the maximum distance threshold and greater than or equal to the minimum distance threshold, the two measured boundaries C are determined to be the two target boundaries.

[0074] In the above scheme, such as Figure 2 As shown, the maximum distance threshold and the minimum distance threshold can be preset, and the actual distance between the two measured boundaries C can be measured. If the actual distance is less than or equal to the maximum distance threshold and greater than or equal to the minimum distance threshold, then the two measured boundaries C can be considered as the target boundary D. In other words, it is determined that the two measured boundaries C can effectively reflect the true boundary, and the next step of defect determination can be carried out based on the two measured boundaries C.

[0075] In some exemplary embodiments, step S04 above further includes:

[0076] Step S042: If the measured distance between the two measured boundaries C in the wafer radial direction is greater than the maximum distance threshold or less than the minimum distance threshold, verify whether one of the two measured boundaries C is a target boundary.

[0077] Step S043: If the verification result is that one of the two measured boundaries C is a target boundary, obtain the other target boundary based on the verified target boundary;

[0078] Step S044: If the verification result is that neither of the two measured boundaries C is the target boundary, a preset custom boundary is used as the target boundary, and based on the custom boundary, the other target boundary is obtained.

[0079] In the above scheme, if the measured distance between the two measured boundaries C in the wafer radial direction is greater than the maximum distance threshold or less than the minimum distance threshold, then the two measured boundaries C are considered to be invalid as they cannot truly reflect the wafer boundary line. In this case, step S042 can be executed to verify the reliability of the two measured boundaries C to determine whether one of the measured boundaries C can be used as the target boundary. If one of the measured boundaries C is reliable, then it can be considered as a target boundary, and the other target boundary can be obtained. If both measured boundaries C are unreliable, then a preset custom boundary can be used as the target boundary.

[0080] For example, the two measured boundaries C may include a first measured boundary C1 and a second measured boundary C2.

[0081] For example, step S042 above may specifically include:

[0082] Step S0421: Offset the first measured boundary C1 by a first distance H1 in the direction closer to the second measured boundary C2 to obtain the first virtual boundary C1';

[0083] Step S0422: If the length ratio of the wafer image corresponding to the first virtual boundary C1' in the region to be identified is greater than or equal to the second ratio threshold, the first measured boundary C1 is determined to be a target boundary D;

[0084] Step S0423: If the length ratio of the wafer image corresponding to the first virtual boundary C1' in the area to be identified is less than the second ratio threshold, the second measured boundary C2 is shifted by the first distance H1 towards the direction closer to the first measured boundary C1 to obtain the second virtual boundary C2';

[0085] Step S0424: If the length ratio of the wafer image corresponding to the second virtual boundary C2' in the region to be identified is greater than or equal to the second ratio threshold, the second measured boundary C2 is determined to be a target boundary D.

[0086] In the above scheme, the reliability of the first measured boundary C1 of the two measured boundaries C is first verified. When the reliability verification of the first measured boundary C1 fails, the reliability verification of the second measured boundary C2 is then performed.

[0087] For specific reliability verification methods, please refer to [link / reference]. Figure 3As shown, the first measured boundary C1 is offset by a first distance H1 towards the second measured boundary C2. The first distance H1 can cover the width of the wafer edge chipping. For example, the first distance H1 can be 4 to 8 distance units (e.g., 1 distance unit can refer to 1 pixel). For instance, the first distance H1 can be 4, 5, 6, 7, or 8 distance units, thus obtaining the first virtual boundary C1'. Next, it is verified whether the length ratio of the wafer image corresponding to the first virtual boundary C1' in the area to be identified is greater than the second ratio threshold. For example, the second ratio threshold can be 60% to 80%, such as 60%, 70%, or 80%. The purpose of this is to offset the first measured boundary C1 towards the second measured boundary C2 by a distance that can cover the width of the edge chipping. If the first virtual boundary C1' can still cover most of the edge, it indicates that the selected first measured boundary C1 has no or very few chipped edges, and the first measured boundary C1 is valid and can be used as a target boundary D. Conversely, the first measured boundary C1 is invalid. Similarly, please refer to... Figure 4 As shown, it can be verified whether the second measured boundary C2 is reliable.

[0088] In some exemplary embodiments, step S043 specifically includes:

[0089] Using one of the verified target boundaries D as a reference, a boundary is obtained by offsetting a second distance H2 in the direction of another measured boundary C closer to its opposite side, which is taken as the other target boundary D.

[0090] In the above scheme, such as Figure 3 As shown, taking the verification result that the first measured boundary C1 is reliable and effective as an example, the first measured boundary C1 can be used as the first target boundary D1. Based on the first measured boundary C1, the second target boundary D2 is obtained by shifting the second distance H2 in the direction closer to the second measured boundary C2.

[0091] Wherein, when the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is within the distance threshold, the second distance H2 is the measured distance of the wafer image in the region to be identified in the radial direction of the wafer.

[0092] For example, the distance threshold can be 290μm to 360μm. When the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is within 290μm to 360μm, it is considered that the measured distance is the distance between the two target boundaries D of the wafer. Therefore, the measured distance can be shifted towards the second target boundary D2 based on the first target boundary D1 to obtain the second target boundary D2.

[0093] Furthermore, when the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is not within the distance threshold, the second distance H2 is a preset distance value, which can be a distance value obtained based on testing or experience.

[0094] For example, the distance threshold can be 290μm to 360μm. When the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is not within 290μm to 360μm, it is considered that the measured distance cannot reflect the distance between the two target boundaries D of the wafer. Therefore, based on the first target boundary D1, a distance value obtained by testing or experience can be shifted towards the second target boundary D2 to obtain the second target boundary D2.

[0095] Furthermore, in some exemplary embodiments, step S044 specifically includes: as follows Figure 5 As shown, a boundary is obtained by offsetting the custom boundary E by a second distance H2 in the direction of another measured boundary C on the opposite side, and this boundary is used as the other target boundary D.

[0096] In the above scheme, if both measured boundaries C are determined to be invalid, a default custom boundary E can be used as the first target boundary D1, and then the first target boundary D1 can be offset by a second distance H2 in the direction closer to the other measured boundary C to obtain the second target boundary D2.

[0097] Wherein, when the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is within the distance threshold, the second distance H2 is the measured distance of the wafer image in the region to be identified in the radial direction of the wafer.

[0098] For example, the distance threshold can be 290μm to 360μm. When the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is within 290μm to 360μm, it is considered that the measured distance is the distance between the two target boundaries D of the wafer. Therefore, the measured distance can be shifted towards the second target boundary D2 based on the first target boundary D1 to obtain the second target boundary D2.

[0099] Furthermore, when the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is not within the distance threshold, the second distance H2 is a preset distance value, which can be a distance value obtained based on testing or experience.

[0100] For example, the distance threshold can be 290μm to 360μm. When the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is not within 290μm to 360μm, it is considered that the measured distance cannot reflect the distance between the two target boundaries D of the wafer. Therefore, based on the first target boundary D1, a distance value obtained by testing or experience can be shifted towards the second target boundary D2 to obtain the second target boundary D2.

[0101] The wafer defect detection method provided by the embodiments of this disclosure can avoid the problem of missing large defects.

[0102] For example, regarding wafer defect detection methods employing related technologies, compared to the wafer defect detection method provided in the embodiments of this disclosure, for example... Figures 6A to 6D When detecting different types of defects, such as chip defects, crack defects, and edge chipping defects, the wafer defect detection methods in related technologies cannot correctly detect the defects, while the wafer defect detection methods provided in the embodiments of this disclosure can correctly detect the defects.

[0103] Specifically, for example, Figure 6A As shown, if the wafer defect detection method in the relevant technology is used, for Figure 6A When the image in the area to be identified is compared in grayscale, the first measured boundary C1 is incorrectly identified as the true boundary of the edge defect due to the large size of the defect. Therefore, the defect type and defect level identified based on the first measured boundary C1 are incorrect. However, the wafer defect detection method of this embodiment can determine the first target boundary D1 based on the first measured boundary C1, and the defect type and defect level identified based on the first target boundary D1 are correct.

[0104] For example, such as Figure 6B As shown, if the wafer defect detection method in the relevant technology is used, for Figure 6B The image in the area to be identified is compared in grayscale. Due to the large edge chipping defects at the edge of the wafer, the true boundary of the edge defect cannot be identified. Therefore, the defect cannot be identified based on the first measured boundary C1. However, the wafer defect detection method of this embodiment can determine the first target boundary D1 based on the first measured boundary C1. The defect type and defect level identified based on the first target boundary D1 are correct.

[0105] For example, Figure 6C As shown, if the wafer defect detection method in the relevant technology is used, for Figure 6CThe image in the area to be identified is compared in grayscale. For crack defects at the edge of the wafer, the true boundary of the edge defect cannot be accurately identified. Therefore, the identified defect type and defect level are incorrect. However, the wafer defect detection method of this embodiment determines the second target boundary D2. The defect type and defect level identified based on the second target boundary D2 are correct.

[0106] Specifically, for example, Figure 6D As shown, if the wafer defect detection method in the relevant technology is used, for Figure 6D When the image in the area to be identified is compared in grayscale, the second measured boundary C2 is incorrectly identified as the true boundary of the edge defect due to the large size of the defect. Therefore, the defect type and defect level identified based on the second measured boundary C2 are incorrect. However, the wafer defect detection method of this embodiment can determine the second target boundary D2 based on the second measured boundary C2, and the defect type and defect level identified based on the second target boundary D2 are correct.

[0107] Furthermore, this disclosure provides a wafer defect detection apparatus, including a processor and a memory; the processor is used to execute instructions stored in the memory to implement the wafer defect detection method described above.

[0108] Obviously, the wafer defect detection device provided in this embodiment also has the beneficial effects of the wafer defect detection method provided in this embodiment, and will not be described in detail here.

[0109] Furthermore, embodiments of this disclosure provide a computer storage medium storing at least one instruction, which is executed by a processor to implement the wafer defect detection method described above.

[0110] The aforementioned computer-readable storage medium, since the computer program stored in its memory implements the steps in the above-described method embodiments when executed by a processor, can similarly have the beneficial effects brought about by the above-described methods.

[0111] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0112] The following points need to be explained:

[0113] (1) The accompanying drawings of the embodiments of this disclosure only involve the structures involved in the embodiments of this disclosure. Other structures can be referred to the general design.

[0114] (2) For clarity, the thickness of layers or regions is enlarged or reduced in the drawings used to describe embodiments of the present disclosure, i.e., these drawings are not drawn to actual scale. It will be understood that when an element such as a layer, film, region or substrate is referred to as being “above” or “below” another element, the element may be “directly” located “above” or “below” the other element or there may be intermediate elements.

[0115] (3) Where there is no conflict, the embodiments of this disclosure and the features in the embodiments can be combined with each other to obtain new embodiments.

[0116] The above are merely specific embodiments of this disclosure, but the scope of protection of this disclosure is not limited thereto. The scope of protection of this disclosure shall be determined by the scope of the claims.

Claims

1. A method for detecting wafer defects, characterized in that, include: Obtain the edge image of the wafer; Based on the image features of the edge image, determine the region to be identified that has defects; Based on the image features in the region to be identified, two measured boundaries of the wafer radial direction in the region to be identified are identified; Based on the two measured boundaries, two target boundaries in the wafer radial direction within the region to be identified are determined; Based on the image features of the two target boundaries and the region to be identified, the defect information within the region to be identified is determined; Specifically, identifying the two measured boundaries of the wafer radial direction in the region to be identified based on image features in the region to be identified includes: In the grayscale image corresponding to the region to be identified, wafer images and non-wafer images with grayscale difference values ​​greater than the grayscale threshold are identified; Based on the morphological features of the wafer image and the non-wafer image, a reference region of a wafer is selected. The two boundaries of the reference region of the wafer have an extension length along the edge of the wafer in the region to be identified that is greater than or equal to a first percentage threshold. The two boundaries of the reference region of the wafer are determined to be the two measured boundaries; The determination of two target boundaries in the wafer radial direction within the region to be identified, based on the two measured boundaries, specifically includes: If the measured distance between the two measured boundaries in the wafer radial direction is less than or equal to the maximum distance threshold and greater than or equal to the minimum distance threshold, the two measured boundaries are determined to be the two target boundaries. The determination of two target boundaries in the wafer radial direction within the region to be identified, based on the two measured boundaries, specifically includes: If the measured distance between the two measured boundaries in the wafer radial direction is greater than the maximum distance threshold or less than the minimum distance threshold, verify whether one of the two measured boundaries is a target boundary. If the verification result shows that one of the two measured boundaries is a target boundary, then based on the verified target boundary, obtain the other target boundary; If the verification result is that neither of the two measured boundaries is the target boundary, a preset custom boundary is used as the target boundary, and the other target boundary is obtained based on the custom boundary. The two measured boundaries include a first measured boundary and a second measured boundary; verifying whether one of the two measured boundaries is a target boundary specifically includes: The first measured boundary is offset by a first distance in a direction closer to the second measured boundary to obtain the first virtual boundary; If the length proportion of the wafer image corresponding to the first virtual boundary in the region to be identified is greater than or equal to the second proportion threshold, the first measured boundary is determined to be a target boundary. If the length proportion of the wafer image corresponding to the first virtual boundary in the region to be identified is less than the second proportion threshold, the second measured boundary is shifted to the direction closer to the first measured boundary by the first distance to obtain the second virtual boundary; If the length proportion of the wafer image corresponding to the second virtual boundary in the region to be identified is greater than or equal to the second proportion threshold, the second measured boundary is determined to be a target boundary.

2. The wafer defect detection method according to claim 1, characterized in that, The step of obtaining another target boundary based on a verified target boundary specifically includes: Using one of the verified target boundaries as a reference, a boundary is obtained by offsetting it by a second distance in the direction of another measured boundary on its opposite side, which is taken as the other target boundary.

3. The wafer defect detection method according to claim 1, characterized in that, The step of obtaining another target boundary based on the custom boundary specifically includes: Using the custom boundary as a reference, a boundary is obtained by offsetting it by a second distance in the direction of another measured boundary on its opposite side, which is used as the other target boundary.

4. The wafer defect detection method according to any one of claims 2 or 3, characterized in that, When the measured distance of the wafer image in the region to be identified along the wafer radial direction is within the distance threshold, the second distance is the measured distance of the wafer image in the region to be identified along the wafer radial direction. When the measured distance of the wafer image in the region to be identified in the radial direction of the wafer is not within the distance threshold, the second distance is a preset distance value.

5. A wafer defect detection device, characterized in that, It includes a processor and a memory; the processor is used to execute instructions stored in the memory to implement the wafer defect detection method as described in any one of claims 1 to 4.

6. A computer storage medium, characterized in that, The storage medium stores at least one instruction, which is executed by a processor to implement the wafer defect detection method as described in any one of claims 1 to 4.

Images