Defect detection methods, devices, equipment and storage media

Abstract

This application discloses a defect detection method, apparatus, device, and storage medium. The method includes: acquiring a measurement image including the edge of a wafer to be inspected; determining an image region to be inspected from the measurement image; extracting features from the image region to obtain pixel distribution features of the image region to be inspected; and performing defect detection on the edge based on the pixel distribution features of the image region to be inspected. In this embodiment, labor costs can be reduced, and the occurrence of missed or false detections can be minimized, allowing for timely detection of defects on the edge of the wafer to be inspected. Furthermore, it can effectively improve the efficiency of defect detection.

Description

Technical Field

[0001] This application relates to, but is not limited to, the semiconductor field, and particularly to a defect detection method, apparatus, device, and storage medium. Background Technology

[0002] In semiconductor component manufacturing processes, defects at the wafer edge (i.e., the wafer edge) have a significant impact on the process and product yield. Related technologies typically use images of the wafer edge taken by metrology equipment after processing (such as photolithography or etching) as measurement images. These images are then manually observed to determine the presence of defects. However, manual defect detection is costly and prone to missed or false positives. Furthermore, the accuracy of defect assessment in manual methods is only assessed during yield testing, typically with a two-week lag, leading to delayed defect detection. Summary of the Invention

[0003] In view of this, embodiments of this application provide a defect detection method, apparatus, device, and storage medium.

[0004] The technical solution of this application embodiment is implemented as follows:

[0005] On one hand, embodiments of this application provide a defect detection method, the method comprising:

[0006] Acquire measurement images including the crystal edges of the wafer to be inspected;

[0007] Determine the image region to be detected from the measurement image;

[0008] Feature extraction is performed on the image region to be detected to obtain the pixel distribution features of the image region to be detected;

[0009] Based on the pixel distribution characteristics of the image region to be detected, defect detection is performed on the crystal edge.

[0010] On the other hand, embodiments of this application provide a defect detection device, the device comprising:

[0011] The first acquisition module is used to acquire a measurement image including the crystal edge of the wafer to be inspected;

[0012] The first determining module is used to determine the image region to be detected from the measurement image;

[0013] The extraction module is used to extract features from the image region to be detected, and obtain the pixel distribution features of the image region to be detected;

[0014] The detection module is used to perform defect detection on the crystal edge based on the pixel distribution characteristics of the image region to be detected.

[0015] In another aspect, embodiments of this application provide a computer device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the program to implement some or all of the steps in the above-described method.

[0016] In another aspect, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements some or all of the steps in the above-described method.

[0017] In this embodiment, firstly, a measurement image including the edge of the wafer to be inspected is acquired; secondly, an image region to be inspected is determined from the measurement image; then, feature extraction is performed on the image region to be inspected to obtain the pixel distribution features of the image region to be inspected; finally, defect detection is performed on the edge based on the pixel distribution features of the image region to be inspected. This reduces labor costs and minimizes missed or false detections, enabling timely detection of defects on the edge of the wafer to be inspected. Furthermore, since edge defect detection is based on the pixel distribution features of the image region to be inspected, and the image region to be inspected is determined from the measurement image of the edge of the wafer to be inspected, the efficiency of defect detection can be effectively improved. Attached Figure Description

[0018] Figure 1 A schematic diagram illustrating the implementation process of a defect detection method provided in an embodiment of this application;

[0019] Figure 2 A schematic diagram illustrating the implementation process of a defect detection method provided in an embodiment of this application;

[0020] Figure 3 A schematic diagram illustrating the implementation process of a defect detection method provided in an embodiment of this application;

[0021] Figure 4 A schematic diagram illustrating the implementation process of a defect detection method provided in an embodiment of this application;

[0022] Figure 5 A schematic diagram illustrating the implementation process of a defect detection method provided in an embodiment of this application;

[0023] Figure 6A This is a schematic diagram of an image region to be detected in a measurement image provided in an embodiment of this application;

[0024] Figure 6BA schematic diagram showing the change in the proportion of white pixels in different height ranges of an image region at different angles within a defect sub-image of an embodiment of this application;

[0025] Figure 6C A schematic diagram comparing image regions with and without crystal edge defects in a defect detection method provided in this application embodiment;

[0026] Figure 6D A schematic diagram showing the change curve of the proportion of white pixels in the image region to be detected in a defective sub-image of an embodiment of this application at different angles;

[0027] Figure 7 This is a schematic diagram of the composition structure of a defect detection device provided in an embodiment of this application;

[0028] Figure 8 This is a schematic diagram of the hardware entity of a computer device provided in an embodiment of this application. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application are further described in detail below with reference to the accompanying drawings and embodiments. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0030] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0031] If the application documents contain similar descriptions such as "first / second", the following explanation shall be added: In the following description, the terms "first / second / third" are used only to distinguish similar objects and do not represent a specific order of objects. It is understood that "first / second / third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used herein is for descriptive purposes only and is not intended to limit the scope of this application.

[0033] This application provides a defect detection method, which can be executed by a processor of a computer device. The computer device can refer to any suitable device with data processing capabilities, such as a server, laptop, tablet, desktop computer, smart TV, set-top box, or mobile device (e.g., mobile phone, portable video player, personal digital assistant, dedicated messaging device, portable gaming device). Figure 1 This is a schematic diagram illustrating the implementation process of a defect detection method provided in an embodiment of this application, as shown below. Figure 1 As shown, the method includes the following steps S101 to S104:

[0034] Step S101: Obtain a measurement image including the edge of the wafer to be inspected.

[0035] Here, the wafer edge to be inspected refers to the edge of the wafer to be inspected, which may include the side surface area of ​​the wafer, the area where the side surface of the wafer is adjacent to the upper surface, and the area where the side surface of the wafer is adjacent to the lower surface.

[0036] The measurement image is an image including the wafer edge, acquired during the wafer fabrication process. In practice, during wafer fabrication, an image acquisition device can be used to capture images of the wafer edge to obtain a measurement image including the wafer edge. The image acquisition device can be installed at any suitable location on the equipment; there are no limitations on this.

[0037] In some embodiments, the measurement image may include the edge of the wafer to be tested, a portion of the upper surface of the wafer to be tested adjacent to the edge, and a portion of the lower surface of the wafer to be tested adjacent to the edge.

[0038] Step S102: Determine the image region to be detected from the measurement image.

[0039] Here, the image region to be detected is the image region in the measurement image where the crystal edge defect features are more obvious. In this image region, crystal edge defects can be detected more accurately, thereby improving the accuracy of crystal edge defect detection.

[0040] In implementation, the image region to be detected can be determined from the measurement image based on set location information, or it can be determined based on the location of image regions with obvious crystal edge defect characteristics in historical measurement images. Alternatively, a trained neural network model can be used to automatically identify the image region to be detected in the measurement image. Those skilled in the art can use any suitable method to determine the image region to be detected from the measurement image according to the actual situation, and the embodiments of this application are not limited in this regard.

[0041] Step S103: Extract features from the image region to be detected to obtain the pixel distribution features of the image region to be detected.

[0042] Here, the pixel distribution features of the image region to be detected can include any suitable features that can reflect the distribution of different pixels in the image region to be detected.

[0043] In some embodiments, the image region to be detected can be quantified, and the distribution of pixels with different values ​​in the image region can be statistically analyzed to obtain the pixel distribution features of the image region to be detected. Here, the pixel distribution features of the image region to be detected may include, but are not limited to, one or more of the following: the total number of pixels corresponding to at least one value in the image region to be detected, the proportion of pixels corresponding to at least one value, and the distribution status of pixels corresponding to at least one value at different positions in the image region to be detected. For example, the image region to be detected can be binarized, and the distribution of black and white pixels in the binarized image region to be detected can be statistically analyzed to obtain the total number of black / white pixels in the image region to be detected, the difference between the number of black and white pixels, the ratio of the number of black and white pixels, the proportion of black / white pixels, and the distribution of black / white pixels at different positions in the image region to be detected.

[0044] Step S104: Based on the pixel distribution characteristics of the image region to be detected, perform defect detection on the crystal edge.

[0045] Here, compared to the case where there are no defects at the crystal edge, the distribution of pixels in the image region to be detected will change when there are defects at the crystal edge. Therefore, based on the pixel distribution characteristics of the image region to be detected, it is possible to detect whether there are defects in the image region to be detected and the distribution of defects.

[0046] In implementation, the pixel distribution characteristics of the image region to be detected can be matched with the pixel distribution characteristics of a pre-defined characterizing image region that are defect-free to determine whether a defect exists at the crystal edge; alternatively, the pixel distribution characteristics of the image region to be detected can be matched with the pixel distribution characteristics of a pre-defined characterizing image region that are defective to determine whether a defect exists at the crystal edge; furthermore, the pixel distribution characteristics of the image region to be detected can be matched with the pixel distribution characteristics of a pre-defined characterizing image region that contain a specific type of defect to determine whether a defect of that specific type exists at the crystal edge. Those skilled in the art can use appropriate methods based on the pixel distribution characteristics of the image region to be detected to perform defect detection on the crystal edge according to the actual situation, and the embodiments of this application are not limited in this regard.

[0047] In some embodiments, step S102 above may include:

[0048] Step S111: Based on preset features, determine the image region to be detected from the measurement image.

[0049] Here, the preset feature can be any suitable feature that can be used to identify image regions with relatively obvious crystal edge defect characteristics. In implementation, the preset feature can be one or more of the following, including but not limited to preset location ranges, pixel distribution features, and crystal edge features. For example, if the preset feature includes a preset location range, the image region in the measurement image corresponding to that location range can be determined as the image region to be detected. Similarly, if the preset feature includes a preset pixel distribution feature, the image region in the measurement image containing that pixel distribution feature can be determined as the image region to be detected.

[0050] In some embodiments, the method may further include:

[0051] Step S121: If it is determined that there is a defect in the crystal edge, generate and send a warning message.

[0052] Here, the warning information refers to information used to alert the user to defects on the edge of the wafer under inspection. This information may include, but is not limited to, one or more of the following: voice warnings, warning indicator lights, warning phone calls, warning emails, and instant messaging software messages. Users can take appropriate warning responses based on the warning information. For example, upon receiving a warning, they can stop the operation of the relevant process chamber, locate the cause of the defect, or perform equipment maintenance.

[0053] In some embodiments, the method may further include the following steps S131 to S132:

[0054] Step S131: Determine the process chamber corresponding to the measurement image.

[0055] Here, wafer processing is typically completed in at least one process chamber. Different processing steps take place in different process chambers, and therefore, the measurement images acquired during different wafer processing steps will correspond to different process chambers. In practice, when acquiring measurement images, the images can be mapped to the corresponding process chambers, thus allowing the process chamber corresponding to a given measurement image to be determined.

[0056] Step S132: If it is determined that there is a defect in the crystal edge, stop the operation of the machine in the process chamber.

[0057] In some embodiments, the method may further include the following steps S141 to S142:

[0058] Step S141: In response to the data query operation on the crystal edge anomaly trend query interface, obtain the time range and process chamber to be queried.

[0059] Here, the crystal edge anomaly trend query interface can be any suitable interface running on the terminal device used to query crystal edge anomaly trends. Users can perform data query operations on the crystal edge anomaly trend query interface.

[0060] The time range and process chamber to be queried can be preset or entered by the user on the crystal edge anomaly trend query interface; there are no restrictions here.

[0061] Step S142: Query the pixel distribution characteristics of the image region to be detected in each measurement image collected within the time range and corresponding to the process chamber.

[0062] Here, after obtaining the pixel distribution features of the image region to be detected in each measurement image, the acquisition time of the measurement image, the corresponding process chamber, and the pixel distribution features of the image region to be detected in the measurement image can be associated and stored. Based on the time range and process chamber to be queried, the pixel distribution features of the image region to be detected in each measurement image acquired within that time range and corresponding to that process chamber can be retrieved.

[0063] Step S143: The pixel distribution characteristics of the image region to be detected in each measurement image are displayed on the crystal edge anomaly trend query interface.

[0064] Here, the pixel distribution characteristics of the image region to be detected in the measurement image can be displayed in any suitable way on the crystal edge anomaly trend query interface, depending on the actual situation. For example, the pixel distribution characteristics of the image region to be detected in each measurement image can be displayed in the form of a data table, trend chart, or bar chart.

[0065] By displaying the pixel distribution characteristics of the image region to be detected in each measurement image corresponding to the process chamber to be queried within the time range to be queried in the crystal edge anomaly trend query interface, the pixel distribution characteristics of the image region to be detected in the measurement images of different process processing can be visually observed, thereby providing an intuitive judgment or reference for the risk of crystal edge defects in different process processing.

[0066] In this embodiment, firstly, a measurement image including the edge of the wafer to be inspected is acquired; secondly, an image region to be inspected is determined from the measurement image; then, feature extraction is performed on the image region to be inspected to obtain the pixel distribution features of the image region to be inspected; finally, defect detection is performed on the edge based on the pixel distribution features of the image region to be inspected. This reduces labor costs and minimizes missed or false detections, enabling timely detection of defects on the edge of the wafer to be inspected. Furthermore, since edge defect detection is based on the pixel distribution features of the image region to be inspected, and the image region to be inspected is determined from the measurement image of the edge of the wafer to be inspected, the efficiency of defect detection can be effectively improved.

[0067] This application provides a defect detection method, which can be executed by the processor of a computer device. For example... Figure 2 As shown, the method includes the following steps S201 to S206:

[0068] Step S201: Obtain a preset abnormal image library; wherein the abnormal image library includes at least one abnormal measurement image, and each abnormal measurement image contains a defect.

[0069] Here, each anomalous measurement image in the anomalous image library may contain at least one defect. The anomalous image library may be pre-determined based on historical measurement images, obtained from the Internet, or automatically generated using image processing technology. In implementation, those skilled in the art can use appropriate methods to obtain the pre-set anomalous image library according to actual needs, and the embodiments of this application are not limited in this regard.

[0070] Step S202: Determine preset features based on the image region where the defect is located in each of the abnormal measurement images.

[0071] Here, any suitable feature in the image region where the defect is located in each abnormal measurement image can be extracted as a preset feature. In implementation, the location information of the image region where the defect is located in each abnormal measurement image can be analyzed, and the obtained location features can be used as preset features; alternatively, the pixel distribution of the image region where the defect is located in each abnormal measurement image can be analyzed, and the obtained pixel distribution features can be used as preset features; furthermore, the image features of the image region where the defect is located in each abnormal measurement image can be analyzed, and the obtained image features can be used as preset features.

[0072] Step S203: Obtain a measurement image including the edge of the wafer to be inspected.

[0073] Step S204: Based on preset features, determine the image region to be detected from the measurement image.

[0074] Step S205: Extract features from the image region to be detected to obtain the pixel distribution features of the image region to be detected.

[0075] Step S206: Based on the pixel distribution characteristics of the image region to be detected, perform defect detection on the crystal edge.

[0076] Here, steps S203 to S206 correspond to the aforementioned steps S101, S111, S103 and S104. In implementation, the specific implementation methods of the aforementioned steps S101, S111, S103 and S104 can be referred to.

[0077] It should be noted that the implementation of steps S201 to S206 is not limited to the following: Figure 2 The execution order shown can also be such that steps S201 and S202 are executed after step S203.

[0078] In some embodiments, the preset feature includes a preset position range in the vertical dimension, and step S202 may include the following steps S211 to S212:

[0079] Step S211: Determine the position of the image region where the defect is located in each of the abnormal measurement images in the vertical dimension of the corresponding abnormal measurement image.

[0080] Here, the vertical dimension is the directional dimension in the measurement image or abnormal measurement image that corresponds to the measurement direction of the wafer thickness.

[0081] Step S212: Based on the position of the image region where each defect is located in the vertical dimension of the corresponding abnormal measurement image, determine the preset position range in the vertical dimension.

[0082] Here, the preset location range in the vertical dimension refers to the range of locations in the measurement image where the probability of finding crystal edge defects is relatively high. In practice, the location of each defect's image region in the corresponding abnormal measurement image in the vertical dimension can be analyzed to determine the distribution of defects at different locations in the vertical dimension, thereby determining the preset location range in the vertical dimension. For example, the range with the highest number of defects in the vertical dimension can be determined as the preset location range in the vertical dimension, or the range with the highest defect distribution density in the vertical dimension can be determined as the preset location range in the vertical dimension.

[0083] In some embodiments, the preset feature includes a first preset pixel distribution feature in the horizontal dimension, and step S202 may include the following steps S221 to S223:

[0084] Step S221: Based on the image region where the defect is located in each of the abnormal measurement images, determine a first image region set; wherein each image region in the first image region set includes at least one defect.

[0085] Here, the first image region set may include the image region where the defect is located in each abnormal measurement image.

[0086] Step S222: Determine the pixel distribution characteristics of each image region in the first image region set in the horizontal dimension.

[0087] Here, the horizontal dimension is the directional dimension perpendicular to the vertical dimension in the measured image. The pixel distribution characteristics of the image region in the horizontal dimension may include, but are not limited to, one or more of the following: the number of pixels corresponding to at least one value in the image region at different positions in the horizontal dimension after numerical processing; the proportion of pixels corresponding to at least one value in the image region at different positions in the horizontal dimension; and the distribution state of pixels corresponding to at least one value in the image region at different positions in the horizontal dimension.

[0088] In practice, the image region can be quantified, and the distribution of pixels with different values ​​in the horizontal dimension of the quantified image region can be analyzed to obtain the pixel distribution characteristics of the image region in the horizontal dimension. For example, the image region can be binarized, and the distribution of black and white pixels in the horizontal dimension of the binarized image region can be statistically analyzed to obtain the total number of black / white pixels in the image region, the difference between the number of black and white pixels, the ratio of the number of black and white pixels, the proportion of black / white pixels, and the distribution of black / white pixels at different positions in the horizontal dimension of the image region.

[0089] Step S223: Determine the first preset pixel distribution features based on the pixel distribution features of each image region in the horizontal dimension.

[0090] Here, the first preset pixel distribution feature is the pixel distribution feature in the horizontal dimension of the image region where the crystal edge defect features are relatively obvious in the measurement image. In implementation, any suitable feature extraction method can be used to extract the first preset pixel distribution feature from the pixel distribution feature in the horizontal dimension of each image region, or statistical analysis can be performed on the pixel distribution feature in the horizontal dimension of each image region to obtain the first preset pixel distribution feature; there is no limitation here.

[0091] In some embodiments, step S221 may include: step S2211, cutting each of the abnormal measurement images in the vertical dimension to obtain a second image region set; step S2212, based on the image region where the defect is located in each of the abnormal measurement images, selecting image regions including at least one defect from the second image region set to obtain a first image region set. Here, each image region in the second image region set corresponds to a different position range in the vertical dimension. In some embodiments, each abnormal measurement image may be equally divided in the vertical dimension to obtain the second image region set.

[0092] In some embodiments, step S204 may include the following steps S231 to S232:

[0093] Step S231: The measurement image of the crystal edge of the wafer to be tested is cut in the vertical dimension to obtain a third image region set;

[0094] Step S232: Based on the first preset pixel distribution features, determine the image region to be detected from the third image region set.

[0095] Here, the image region in the third image region that matches the first preset pixel distribution features can be determined as the image region to be detected.

[0096] In some embodiments, the preset feature includes a preset crystal edge feature, and the image region to be detected includes an image region containing the preset crystal edge feature. Step S204 may include: step S241, performing feature recognition on the measurement image to obtain an image region containing the preset crystal edge feature. Here, the preset crystal edge feature can be any suitable preset feature for identifying crystal edges, and is not limited thereto. In implementation, any suitable image recognition technology can be used to perform feature recognition on the measurement image to obtain an image region containing the preset crystal edge feature.

[0097] In this embodiment, based on the image region where the defect is located in each abnormal measurement image in a preset abnormal image library, preset features are determined, and based on these preset features, the image region to be detected is determined from the measurement image. This allows for the rapid and accurate acquisition of preset features, thereby accurately determining the image region to be detected with relatively obvious crystal edge defect features from the measurement image, and further improving the accuracy of defect detection on crystal edges.

[0098] This application provides a defect detection method, which can be executed by the processor of a computer device. For example... Figure 3 As shown, the method includes the following steps S301 to S304:

[0099] Step S301: Obtain a measurement image including the edge of the wafer to be inspected.

[0100] Step S302: Determine the image region to be detected from the measurement image.

[0101] Step S303: Extract features from the image region to be detected to obtain the pixel distribution features of the image region to be detected.

[0102] Here, steps S301 to S303 correspond to the aforementioned steps S101 to S103, and can be implemented with reference to the specific implementation of the aforementioned steps S101 to S103.

[0103] Step S304: If the pixel distribution characteristics of the image region to be detected do not match the second preset pixel distribution characteristics, it is determined that there is a defect in the crystal edge; wherein, the second preset pixel distribution characteristics indicate that there is no defect in the image region.

[0104] Here, the second preset pixel distribution feature can be any suitable feature used to characterize the absence of defects in the image region, and is not limited here. In implementation, the second preset pixel distribution feature may include, but is not limited to, one or more of the following: the total number of pixels corresponding to at least one value in the numerically converted defect-free image region; the proportion of pixels corresponding to at least one value; and the distribution state of pixels corresponding to at least one value at different locations in the image region. For example, the distribution of black and white pixels in the binarized defect-free image region can be statistically analyzed to obtain the total number of black / white pixels in the defect-free image region, the difference between the number of black and white pixels, the ratio of the number of black and white pixels, the proportion of black / white pixels, and the distribution of black / white pixels at different locations in the defect-free image region.

[0105] In some embodiments, the pixel distribution features of the image region to be detected include the distribution features of black and white pixels in the image region to be detected, and step S303 may include the following steps S311 to S312:

[0106] Step S311: Binarize the image region to be detected.

[0107] Step S312: Determine the distribution characteristics of black and white pixels in the image region to be detected after binarization.

[0108] Here, the distribution characteristics of black and white pixels may include, but are not limited to, one or more of the following: the total number of black pixels / white pixels, the difference between the number of black pixels and white pixels, the ratio of the number of black pixels to white pixels, the ratio of the number of white pixels to black pixels, the proportion of black pixels / white pixels, and the distribution of black pixels / white pixels at different locations in the image area.

[0109] In some embodiments, the second preset pixel distribution feature includes a preset black and white pixel distribution feature, and step S304 may include: step S321, determining that the crystal edge has a defect if it is determined that the distribution feature of black and white pixels in the image region to be detected does not match the preset black and white pixel distribution feature. Here, the preset black and white pixel distribution feature may include, but is not limited to, one or more of the following: a preset threshold for the total number of black pixels / white pixels, a threshold for the difference in the number of black pixels and white pixels, a threshold for the ratio of the number of black pixels to white pixels, a threshold for the ratio of the number of white pixels to black pixels, a threshold for the proportion of black pixels / white pixels, and the distribution of black pixels / white pixels at different locations in the image region.

[0110] In some embodiments, the distribution characteristics of black and white pixels in the image region to be detected include the proportion of white pixels in the image region to be detected, and the preset black and white pixel distribution characteristics include a preset white pixel proportion threshold. Step S321 may include: Step S331, if it is determined that the proportion of white pixels in the image region to be detected is greater than the white pixel proportion threshold, it is determined that there is a defect in the crystal edge.

[0111] In this embodiment, if the pixel distribution characteristics of the image region to be detected do not match the second preset pixel distribution characteristics representing a defect-free image region, a defect is determined to exist at the crystal edge. This allows for a simple and quick determination of whether a defect exists at the crystal edge of the wafer to be inspected.

[0112] This application provides a defect detection method, which can be executed by the processor of a computer device. For example... Figure 4 As shown, the method includes the following steps S401 to S407:

[0113] Step S401: Obtain a measurement image including the edge of the wafer to be inspected.

[0114] Step S402: Determine the image region to be detected from the measurement image.

[0115] Step S403: Extract features from the image region to be detected to obtain the pixel distribution features of the image region to be detected.

[0116] Here, steps S401 to S403 correspond to the aforementioned steps S101 to S103, and can be implemented with reference to the specific implementation of the aforementioned steps S101 to S103.

[0117] Step S404: Obtain a preset normal image library; wherein the normal image library includes at least one defect-free crystal edge image.

[0118] Here, the normal image library can be pre-determined based on historical measurement images, obtained from the Internet, or automatically generated using image processing technology. In implementation, those skilled in the art can use appropriate methods to obtain the pre-set normal image library according to actual needs, and the embodiments of this application are not limited in this regard.

[0119] Step S405: Determine the pixel distribution characteristics in each of the defect-free crystal edge images.

[0120] Here, the pixel distribution features in a defect-free crystal edge image may include, but are not limited to, the pixel distribution features in the horizontal dimension, the pixel distribution features in the vertical dimension, or the pixel distribution features in a two-dimensional coordinate system formed by the horizontal and vertical dimensions.

[0121] Step S406: Determine the second preset pixel distribution features based on the pixel distribution features in each of the defect-free crystal edge images.

[0122] Here, the second preset pixel distribution feature is the pixel distribution feature that characterizes the defect-free region of the image. In implementation, any suitable feature extraction method can be used to extract the second preset pixel distribution feature from the pixel distribution feature of each defect-free crystal edge image, or statistical analysis can be performed on the pixel distribution feature of each defect-free crystal edge image to obtain the second preset pixel distribution feature; there is no limitation here.

[0123] Step S407: If the pixel distribution characteristics of the image region to be detected do not match the second preset pixel distribution characteristics, it is determined that there is a defect in the crystal edge; wherein, the second preset pixel distribution characteristics indicate that there is no defect in the image region.

[0124] Here, step S407 corresponds to the aforementioned step S304, and the specific implementation of the aforementioned step S304 can be referred to during implementation.

[0125] In this embodiment, a second preset pixel distribution feature is determined based on the pixel distribution features in each defect-free crystal edge image from a preset normal image library. This allows for the rapid and accurate acquisition of the second preset pixel distribution feature, thereby improving the accuracy of defect detection on the crystal edge.

[0126] This application provides a defect detection method, which can be executed by the processor of a computer device. For example... Figure 5 As shown, the method includes the following steps S501 to S505:

[0127] Step S501: Obtain a measurement image including the edge of the wafer to be inspected; wherein the width of the measurement image in the horizontal dimension is the same as the perimeter of the wafer to be inspected.

[0128] Here, the measurement image can be an image acquired around the side of the wafer to be tested, and the width of the measurement image in the horizontal dimension is the same as the circumference of the wafer to be tested.

[0129] Step S502: Divide the measurement image into equal parts in the horizontal dimension to obtain multiple sub-images.

[0130] Here, after the measurement image is divided into multiple sub-images in the horizontal dimension, each sub-image can correspond to the same width in the horizontal dimension.

[0131] Step S503: Determine an image region to be detected from each of the sub-images.

[0132] Step S504: Extract features from the image region to be detected to obtain the pixel distribution features of the image region to be detected.

[0133] Step S505: Based on the pixel distribution characteristics of the image region to be detected, perform defect detection on the crystal edge.

[0134] Here, steps S503 to S505 correspond to the aforementioned steps S102 to S104, and in practice, the specific implementation of the aforementioned steps S102 to S104 can be referred to.

[0135] In this embodiment, the width of the measurement image in the horizontal dimension is the same as the circumference of the wafer to be inspected. By dividing the measurement image into equal parts in the horizontal dimension, multiple sub-images are obtained, and an image region to be inspected is determined from each sub-image. In this way, more refined defect detection can be performed on the measurement image, thereby further improving the accuracy of defect detection on the wafer edge.

[0136] This application provides a defect detection method, which includes the following steps S601 to S605:

[0137] Step S601: Based on the abnormal image library, determine the image region to be detected in the measurement image containing the crystal edge of the wafer to be detected.

[0138] Here, the image region to be detected is the region in the measurement image where the crystal edge features are more prominent, corresponding to the location of the crystal edge in the measurement image. For example, the image region to be detected can be an image region in the measurement image within the range of 147.5 to 147.9 nanometers (nm) from the bottom edge of the measurement image.

[0139] Figure 6A This is a schematic diagram of an image region to be detected in a measurement image, provided as an embodiment of this application. For example... Figure 6AAs shown, the image region 62 to be detected is the region in the measurement image 61 where the crystal edge features 63 are more prominent.

[0140] In some embodiments, the image region to be detected can be achieved through the following steps S611 to S612:

[0141] Step S611: Cut the measurement image into X parts in the horizontal dimension to obtain X sub-images; where X is a positive integer greater than 1.

[0142] Here, the edge of the wafer to be inspected includes a 360-degree range corresponding to the circumference of the wafer's side surface, and each sub-image obtained after cutting can correspond to the 360 / X-degree range of the circumference.

[0143] Step S612: Based on the preset position range in the vertical dimension, determine an image region to be detected from each sub-image.

[0144] Here, the vertical dimension can be the height dimension of the measurement image. The height of the measurement image can be the sum of the height of the crystal edge (i.e., the thickness of the wafer), the height of the region adjacent to the crystal edge on the upper surface of the wafer, and the height of the region adjacent to the crystal edge on the lower surface of the wafer. For example, the height of the measurement image is 5 millimeters (mm). Each sub-image is obtained by cutting the measurement image in the horizontal dimension, so the height of each sub-image is also 5mm. The preset position range in the vertical dimension is the height range where the crystal edge defect characteristics are more obvious in the measurement image, such as the height range of 145nm to 155nm, or the height range of 147.5nm to 147.9nm, etc.

[0145] In some embodiments, each abnormal measurement image in the abnormal image library can be segmented horizontally to obtain multiple sub-images. Each sub-image can then be segmented vertically to obtain multiple image regions corresponding to different locations in the vertical dimension. By analyzing the distribution of black and white pixels in the horizontal dimension of each image region, a preset distribution feature of black and white pixels in the horizontal dimension can be obtained. Based on this preset distribution feature of black and white pixels in the horizontal dimension, the location range where crystal edge defect features are more obvious in the vertical dimension of the measurement image can be determined, which is also the preset location range in the vertical dimension. For example, the proportion of white pixels at different locations in the horizontal dimension of each image region obtained from the abnormal image library can be analyzed. The location range in the vertical dimension corresponding to the image region where the proportion of white pixels at different locations in the horizontal dimension changes significantly or where the average proportion of white pixels is high can be determined as the preset location range in the vertical dimension.

[0146] In some embodiments, a preset position range in the vertical dimension can be intuitively determined from different position ranges in the vertical dimension based on the curve of the proportion change of white pixels in the horizontal dimension in the image region corresponding to different position ranges in the vertical dimension. Figure 6B This is a schematic diagram illustrating the variation of the proportion of white pixels in image regions of different height ranges at different angles within a defect sub-image included in an embodiment of this application. See also... Figure 6B The position of the measurement image in the horizontal dimension can be represented by the angle corresponding to the image region on the wafer, and the position of the measurement image in the vertical dimension can be represented by the height of the measurement image. The sub-image containing defects with an angle range of 185 degrees to 195 degrees is cut in the vertical dimension to obtain multiple image regions, each of which corresponds to a height range. Figure 6B In the graph, the horizontal axis represents angle, and the vertical axis represents the proportion of white pixels in the region corresponding to different angles for each image area. The proportion of white pixels in the region corresponding to each angle for each image area can be represented by a scatter point in the coordinate system (e.g., scatter point 64). The scatter points corresponding to image areas within the same height range can be fitted to a white pixel proportion distribution curve (e.g., curve 65). Different curves in the coordinate system can reflect the change in the proportion of white pixels in the region corresponding to a specific height range in a sub-image containing defects at different angles. The curve graph allows for a visual determination of the preset height range in the vertical dimension from different height ranges. For example, the height range corresponding to the curve with the largest maximum value can be determined as the preset height range in the vertical dimension, or the height range corresponding to the curve with the largest average value can be determined as the preset height range in the vertical dimension.

[0147] Step S602: Based on the normal image library, determine the range of white pixels in the binarized defect-free crystal edge image.

[0148] Here, the proportion of white pixels (denoted as W value) after binarization in a defect-free crystal edge image falls within this range, while the proportion of white pixels after binarization in a defective crystal edge image does not fall within this range. For example, the proportion of white pixels in a defect-free crystal edge image after binarization can range from 0% to 15%.

[0149] Figure 6C This is a schematic diagram comparing image regions with and without crystal edge defects in the defect detection method provided in this application embodiment. In image region 67, there are no crystal edge defects, and in image region 68, the proportion of white pixels is 80.84%, which is not within the range of 0 to 15% of the proportion of white pixels in the defect-free crystal edge image. Therefore, it can be determined that there are crystal edge defects in image region 68.

[0150] Step S603: For each image region to be detected, perform binarization processing on the image region to be detected to obtain the proportion of white pixels in the image region to be detected after binarization; if the proportion of white pixels is not within the range of the proportion of white pixels in the binarized defect-free crystal edge image, determine that there is a crystal edge defect in the sub-image corresponding to the image region to be detected.

[0151] Step S604: Using the W value as the observation value, determine whether there are crystal edge defects in the X sub-images corresponding to the wafer to be inspected, and if it is determined that there are crystal edge defects in the wafer, use the early warning system to issue an early warning.

[0152] Step S605: Using the W value as the observed value, obtain the trend map of the W value change in the image region to be detected in the measurement images during different process treatments, so as to provide a reference for the risk judgment of crystal edge anomalies.

[0153] In some embodiments, the proportion range of white pixels in the binarized defect-free crystal edge image can also be determined based on an abnormal image library.

[0154] Figure 6D This is a schematic diagram illustrating the change in the proportion of white pixels in the image region to be detected within a defective sub-image provided in an embodiment of this application at different angles. (See also...) Figure 6D The position of the measurement image in the horizontal dimension can be represented by the angle corresponding to the image region on the wafer, and the position of the measurement image in the vertical dimension can be represented by the height of the measurement image. Image regions 1 to 8 are the image regions to be detected in different sub-images containing defects, and each image region corresponds to an angle range. Figure 6D In the graph, the horizontal axis represents angle, and the vertical axis represents the proportion of white pixels in the region corresponding to different angles for each image area. The proportion of white pixels in the region corresponding to each angle for each image area can be represented by a scatter point in the coordinate system. The scatter points corresponding to the same image area can be fitted into a white pixel proportion distribution curve (such as curve 66). Different curves in the coordinate system can reflect the change in the proportion of white pixels in the region corresponding to different angles for the image area to be detected in different sub-images containing defects. The curve graph can intuitively determine the range of white pixel proportions in the binarized defective crystal edge image, and thus determine the range of white pixel proportions in the binarized defect-free crystal edge image. For example, the minimum value of the white pixel proportion in each curve can be determined as the proportion threshold. The range of white pixel proportions in the binarized defective crystal edge image can be greater than or equal to this proportion threshold, while the range of white pixel proportions in the binarized defect-free crystal edge image can be less than this proportion threshold.

[0155] In this embodiment, the method uses a measurement image containing the edge of the wafer to be inspected for defect detection. Upon detection of edge defects, it provides an anomaly warning for the wafer production line, thereby reducing the impact of edge defects on product yield and improving the lag in detecting edge anomalies through manual observation in related technologies. Furthermore, the method can use the distribution characteristics of black and white pixels in the measurement image as observation values ​​to generate a trend map of the black and white pixel distribution characteristics as reference information for risk assessment of edge anomalies, thus improving the problem of unobservable edge anomaly trends in related technologies.

[0156] Figure 7 This is a schematic diagram of the structural composition of a defect detection device provided in an embodiment of this application, as shown below. Figure 7 As shown, the defect detection device 700 includes: a first acquisition module 710, a first determination module 720, an extraction module 730, and a detection module 740, wherein:

[0157] The first acquisition module 710 is used to acquire a measurement image including the crystal edge of the wafer to be inspected;

[0158] The first determining module 720 is used to determine the image region to be detected from the measurement image;

[0159] Extraction module 730 is used to extract features from the image region to be detected, and obtain the pixel distribution features of the image region to be detected;

[0160] The detection module 740 is used to perform defect detection on the crystal edge based on the pixel distribution characteristics of the image region to be detected.

[0161] In some embodiments, the first determining module is further configured to: determine the image region to be detected from the measurement image based on preset features.

[0162] In some embodiments, the apparatus further includes: a second acquisition module, configured to acquire a preset abnormal image library; wherein the abnormal image library includes at least one abnormal measurement image, and each abnormal measurement image contains a defect; and a second determination module, configured to determine the preset feature based on the image region where the defect is located in each abnormal measurement image.

[0163] In some embodiments, the preset feature includes a preset position range in the vertical dimension, and the second determining module is further configured to: determine the position of the image region where the defect is located in each of the abnormal measurement images in the vertical dimension of the corresponding abnormal measurement image; and determine the preset position range in the vertical dimension based on the position of the image region where the defect is located in the corresponding abnormal measurement image in the vertical dimension.

[0164] In some embodiments, the preset feature includes a first preset pixel distribution feature in the horizontal dimension, and the second determining module is further configured to: determine a first image region set based on the image region where the defect is located in each of the abnormal measurement images; wherein each image region in the first image region set includes at least one defect; determine the pixel distribution feature in the horizontal dimension of each image region in the first image region set; and determine the first preset pixel distribution feature based on the pixel distribution feature in the horizontal dimension of each of the image regions.

[0165] In some embodiments, the second determining module is further configured to: cut each of the abnormal measurement images in the vertical dimension to obtain a second image region set; and based on the image region where the defect is located in each of the abnormal measurement images, filter image regions including at least one defect from the second image region set to obtain the first image region set.

[0166] In some embodiments, the first determining module is further configured to: cut the measurement image of the wafer edge to be detected in the longitudinal dimension to obtain a third image region set; and determine the image region to be detected from the third image region set based on the first preset pixel distribution features.

[0167] In some embodiments, the preset feature includes a preset crystal edge feature, and the image region to be detected includes an image region containing the preset crystal edge feature. The first determining module is further configured to: perform feature recognition on the measurement image to obtain an image region containing the preset crystal edge feature.

[0168] In some embodiments, the detection module is further configured to: determine that there is a defect in the crystal edge when it is determined that the pixel distribution characteristics of the image region to be detected do not match the second preset pixel distribution characteristics; wherein the second preset pixel distribution characteristics indicate that there are no defects in the image region.

[0169] In some embodiments, the apparatus further includes: a third acquisition module for acquiring a preset normal image library; wherein the normal image library includes at least one defect-free crystal edge image; a third determination module for determining pixel distribution features in each defect-free crystal edge image; and a fourth determination module for determining a second preset pixel distribution feature based on the pixel distribution features in each defect-free crystal edge image.

[0170] In some embodiments, the pixel distribution features of the image region to be detected include the distribution features of black and white pixels in the image region to be detected, and the extraction module is further configured to: perform binarization processing on the image region to be detected; and determine the distribution features of black and white pixels in the image region to be detected after binarization processing.

[0171] In some embodiments, the second preset pixel distribution feature includes a preset black and white pixel distribution feature, and the detection module is further configured to: determine that there is a defect in the crystal edge when it is determined that the distribution feature of black and white pixels in the image region to be detected does not match the preset black and white pixel distribution feature.

[0172] In some embodiments, the distribution characteristics of black and white pixels in the image region to be detected include the proportion of white pixels in the image region to be detected, and the preset black and white pixel distribution characteristics include a preset white pixel proportion threshold. The detection module is further configured to: determine that there is a defect in the crystal edge when it is determined that the proportion of white pixels in the image region to be detected is greater than the white pixel proportion threshold.

[0173] In some embodiments, the width of the measurement image in the horizontal dimension is the same as the circumference of the wafer to be inspected, and the first determining module is further configured to: divide the measurement image into equal parts in the horizontal dimension to obtain a plurality of sub-images; and determine an image region to be inspected from each of the sub-images.

[0174] In some embodiments, the apparatus further includes: an early warning device for generating and sending early warning information when it is determined that a defect exists on the crystal edge.

[0175] In some embodiments, the apparatus further includes: a fifth determining module for determining a process chamber corresponding to the measurement image; and a stopping module for stopping the operation of the machine in the process chamber when it is determined that there is a defect in the crystal edge.

[0176] In some embodiments, the apparatus further includes: a third acquisition module, configured to acquire a time range and process chamber to be queried in response to a data query operation on a crystal edge anomaly trend query interface; a query module, configured to query the pixel distribution characteristics of the image region to be detected in each measurement image collected within the time range and corresponding to the process chamber; and a display module, configured to display the pixel distribution characteristics of the image region to be detected in each measurement image on the crystal edge anomaly trend query interface.

[0177] The descriptions of the above device embodiments are similar to those of the above method embodiments, and have similar beneficial effects. For technical details not disclosed in the device embodiments of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0178] It should be noted that, in the embodiments of this application, if the above-described defect detection method is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, or the part that contributes to the related technology, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), magnetic disks, or optical disks. Thus, the embodiments of this application are not limited to any specific hardware and software combination.

[0179] Correspondingly, embodiments of this application provide a computer device, including a memory and a processor. The memory stores a computer program that can run on the processor, and the processor executes the program to implement the steps in the above-described method.

[0180] Correspondingly, embodiments of this application provide a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps in the above-described method.

[0181] Correspondingly, embodiments of this application provide a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. When the computer program is read and executed by a computer, it implements some or all of the steps in the above-described method. This computer program product can be implemented specifically through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied as a computer storage medium; in another optional embodiment, the computer program product is specifically embodied as a software product, such as a software development kit (SDK), etc.

[0182] It should be noted that the descriptions of the above-described storage media, computer program products, and device embodiments are similar to the descriptions of the above-described method embodiments, and have similar beneficial effects. For technical details not disclosed in the embodiments of the storage media, computer program products, and devices of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0183] It should be noted that, Figure 8 This is a schematic diagram of a hardware entity of a computer device in an embodiment of this application, such as... Figure 8 As shown, the hardware entity of the computer device 800 includes: a processor 801, a communication interface 802, and a memory 803, wherein:

[0184] Processor 801 typically controls the overall operation of computer device 800.

[0185] The communication interface 802 enables computer devices to communicate with other terminals or servers over a network.

[0186] The memory 803 is configured to store instructions and applications executable by the processor 801, and can also cache data to be processed or already processed by the processor 801 and various modules in the computer device 800 (e.g., image data, audio data, voice communication data and video communication data), which can be implemented by flash memory or random access memory (RAM).

[0187] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0188] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0189] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0190] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0191] In addition, each functional unit in the various embodiments of this application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0192] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0193] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence or the part that contributes to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, magnetic disks, or optical disks.

[0194] The above description is merely an embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A defect detection method, characterized in that, The method includes: Acquire measurement images including the crystal edges of the wafer to be inspected; Determine preset features, wherein the preset features include at least a first preset pixel distribution feature; Based on the preset features, the graphic regions in the measurement image that contain the preset features are determined as the image regions to be detected; The image region to be detected is numerically processed, and the pixels with different values ​​in the image region to be detected are statistically analyzed to obtain the pixel distribution characteristics of the image region to be detected; based on the pixel distribution characteristics of the image region to be detected, defect detection is performed on the crystal edge; The step of determining the first preset pixel distribution features includes: Obtain a preset abnormal image library, wherein the abnormal image library includes at least one abnormal measurement image, and each abnormal measurement image contains a defect; Each of the aforementioned abnormal measurement images is cut along the vertical dimension to obtain a second image region set. Each image region in the second image region set corresponds to a different position range along the vertical dimension. The vertical dimension is the directional dimension in the abnormal measurement image that corresponds to the measurement direction of the wafer thickness. Based on the image region where the defect is located in each of the abnormal measurement images, image regions containing at least one defect are selected from the second image region set to obtain a first image region set; The first image region set is numerically processed, and the distribution of pixels with different values ​​in each image region of the first image region set in the horizontal dimension is statistically analyzed to determine the pixel distribution characteristics of each image region in the first image region set in the horizontal dimension. The horizontal dimension is the direction dimension perpendicular to the vertical dimension in the measurement image. Based on the pixel distribution characteristics of each image region in the first image region set in the horizontal dimension, the first preset pixel distribution characteristics are determined.

2. The method according to claim 1, characterized in that, The preset feature also includes a preset location range, and the step of determining the preset location range includes: Obtain a preset abnormal image library, wherein the abnormal image library includes at least one abnormal measurement image, and each abnormal measurement image contains a defect; Determine the position of the image region where the defect is located in each of the aforementioned abnormal measurement images along the vertical dimension of the corresponding abnormal measurement image; The range of locations with the highest number or density of defects along the vertical dimension is determined as the preset location range.

3. The method according to claim 2, characterized in that, Based on the preset features, the step of determining the graphic region in the measurement image that contains the preset features as the image region to be detected includes: The image region in the measurement image that corresponds to the preset position range is determined as the image region to be detected.

4. The method according to claim 1 or 2, characterized in that, The preset features also include preset crystal edge features.

5. The method according to claim 4, characterized in that, Based on the preset features, the step of determining the graphic region in the measurement image that contains the preset features as the image region to be detected includes: The measurement image is subjected to feature recognition to obtain an image region containing the preset crystal edge features, and the image region containing the preset crystal edge features is determined as the image region to be detected.

6. The method according to claim 1, characterized in that, Based on the first preset pixel distribution features, the step of determining the graphic region in the measurement image containing the preset features as the image region to be detected includes: including: The measurement image of the wafer edge to be tested is cut along the longitudinal dimension to obtain a third image region set; The image regions in the third image region that match the first preset pixel distribution features are determined as the image regions to be detected.

7. The method according to claim 1, characterized in that, The pixel distribution feature is selected from at least one of the following: the total number of black pixels / white pixels in the image region, the difference between the number of black pixels and white pixels, the ratio of the number of black pixels to white pixels, the proportion of black pixels / white pixels, and the distribution of black pixels / white pixels at different locations in the image region.

8. The method according to claim 1, characterized in that, The defect detection of the crystal edge based on the pixel distribution characteristics of the image region to be detected includes: If the pixel distribution characteristics of the image region to be detected do not match the second preset pixel distribution characteristics, it is determined that there is a defect in the crystal edge; wherein, the second preset pixel distribution characteristics indicate that there is no defect in the image region.

9. The method according to claim 8, characterized in that, The method further includes: Obtain a preset normal image library; wherein the normal image library includes at least one defect-free crystal edge image; Determine the pixel distribution characteristics in each of the defect-free crystal edge images; The second preset pixel distribution features are determined based on the pixel distribution features in each defect-free crystal edge image.

10. The method according to claim 8 or 9, characterized in that, The pixel distribution features of the image region to be detected include the distribution features of black and white pixels in the image region to be detected. The step of extracting features from the image region to be detected to obtain the pixel distribution features of the image region to be detected includes: The image region to be detected is binarized; Determine the distribution characteristics of black and white pixels in the image region to be detected after binarization.

11. The method according to claim 10, characterized in that, The second preset pixel distribution feature includes preset black and white pixel distribution features. The step of determining that the crystal edge has a defect when the pixel distribution features of the image region to be detected do not match the second preset pixel features includes: If the distribution characteristics of black and white pixels in the image region to be detected do not match the preset black and white pixel distribution characteristics, it is determined that there is a defect in the crystal edge.

12. The method according to claim 11, characterized in that, The distribution characteristics of black and white pixels in the image region to be detected include the proportion of white pixels in the image region to be detected, and the preset black and white pixel distribution characteristics include a preset white pixel proportion threshold. The step of determining that the crystal edge has a defect when the distribution characteristics of black and white pixels in the image region to be detected do not match the preset black and white pixel distribution characteristics includes: If the proportion of white pixels in the image region to be detected is greater than the white pixel proportion threshold, it is determined that there is a defect in the crystal edge.

13. The method according to any one of claims 1, 6 to 9, characterized in that, The width of the measurement image in the horizontal dimension is the same as the circumference of the wafer to be inspected. Determining the image region to be inspected from the measurement image includes: The measurement image is divided into multiple sub-images by dividing it into equal parts in the horizontal dimension; From each of the sub-images, a region of image to be detected is determined.

14. The method according to any one of claims 1, 6 to 9, characterized in that, The method further includes: If a defect is found in the crystal edge, an early warning message is generated and sent.

15. The method according to any one of claims 1, 6 to 9, characterized in that, The method further includes: Determine the process chamber corresponding to the measurement image; If a defect is found in the crystal edge, the operation of the machine in the process chamber shall be stopped.

16. The method according to any one of claims 1, 6 to 9, characterized in that, The method further includes: In response to the data query operation on the crystal edge anomaly trend query interface, obtain the time range and process chamber to be queried; Query the pixel distribution characteristics of the image region to be detected in each of the measurement images collected within the time range and corresponding to the process chamber; The pixel distribution characteristics of the image region to be detected in each measurement image are displayed on the crystal edge anomaly trend query interface.

17. A defect detection apparatus for implementing the method according to any one of claims 1 to 16, characterized in that, include: The first acquisition module is used to acquire a measurement image including the crystal edge of the wafer to be inspected; The first determining module is used to determine preset features and to determine the image region to be detected from the measurement image; The extraction module is used to perform numerical processing on the image region to be detected, and to count the pixels with different values ​​in the image region to be detected, so as to obtain the pixel distribution characteristics of the image region to be detected. The detection module is used to perform defect detection on the crystal edge based on the pixel distribution characteristics of the image region to be detected.

18. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that can run on the processor, the processor executing the program to implement the steps of the method according to any one of claims 1 to 16.

19. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program performs the steps of the method according to any one of claims 1 to 16.

Images