Wafer defect detection method and device, computer equipment and medium

Abstract

The invention relates to a wafer defect detection method and device, computer equipment and a medium. The method comprises the following steps: acquiring an image of a wafer to be detected; generating a difference image according to each pixel in the to-be-detected wafer image and the gray value of the corresponding neighborhood position; performing gray binarization processing on the difference image according to a preset threshold image to generate a contour binary image; the threshold value graph is obtained by processing according to a minimum value graph after the minimum value graph is generated according to each pixel in the set template graph and a gray value of a corresponding neighborhood position; extracting boundary contour points according to the contour binary image to obtain a defect contour point set; performing defect detection on the defect contour point set according to a preset minimum value directional diagram; the minimum value directional diagram is obtained by processing according to the minimum value diagram. And compared with defect point judgment by using a single threshold, the influence of image quality on detection can be reduced, and the detection accuracy is improved.

Description

Technical Field

[0001] The present application relates to the field of image detection technology, and in particular to a wafer defect detection method, device, computer equipment and medium. Background Art

[0002] With the continuous development and progress of the semiconductor industry, its processing technology requirements are increasing day by day, and semiconductor packaging and testing equipment and technology are also changing accordingly, and the accuracy requirements of the equipment are also increasing. Wafer defect detection is an inevitable process in the semiconductor packaging and testing process. In the traditional wafer defect detection process, the defect judgment is based on the gray value difference between the gray value of the wafer pixel and the golden template, that is, a single threshold is used to determine the defect point. The defect detection rate is seriously affected by the image quality. The traditional wafer defect detection method has the disadvantage of low detection accuracy. Summary of the invention

[0003] Based on this, it is necessary to provide a wafer defect detection method, device, computer equipment and medium that can improve detection accuracy in response to the above problems.

[0004] A first aspect of the present application provides a wafer defect detection method, comprising:

[0005] Acquire an image of a wafer to be tested;

[0006] Generate a difference map according to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested;

[0007] The difference image is subjected to grayscale binarization processing according to a preset threshold image to generate a contour binary image; the threshold image is obtained by generating a minimum value image according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the set template image, and then processing according to the minimum value image;

[0008] Extract boundary contour points according to the contour binary image to obtain a defect contour point set;

[0009] Defect detection is performed on the defect contour point set according to a preset minimum value direction map; the minimum value direction map is obtained by processing according to the minimum value map.

[0010] In one embodiment, before performing grayscale binarization processing on the difference image according to a preset threshold image to generate a contour binary image, the method further includes:

[0011] Traversing each pixel in the set template image, determining the neighborhood of each pixel according to a preset spacing step, and dividing the neighborhood into multiple neighborhood directions;

[0012] Generate a minimum value map according to the grayscale values ​​of the pixels and the neighborhood positions in the corresponding neighborhood directions;

[0013] The minimum value map is processed to obtain a threshold value map and a minimum value direction map.

[0014] In one embodiment, generating a minimum value map according to the grayscale values ​​of the pixels and the neighboring positions in the corresponding neighboring directions includes:

[0015] According to the grayscale values ​​of the pixels and the neighboring positions in the corresponding neighboring directions, the absolute values ​​of the grayscale differences of the corresponding pixels in different neighboring directions are calculated;

[0016] From the absolute values ​​of the maximum grayscale differences of each pixel in different neighborhood directions, the minimum value is selected as the value of the corresponding pixel, and a minimum value map is generated according to the values ​​of each pixel.

[0017] In one embodiment, the processing of the minimum value map to obtain a threshold value map and a minimum value direction map includes:

[0018] Multiply the value of each pixel in the minimum value map by a preset standard deviation threshold, and then perform Gaussian blur processing, and set the grayscale value of pixels less than a set constant to the set constant, to obtain a threshold map;

[0019] From the absolute values ​​of the maximum grayscale differences of each pixel in different neighborhood directions in the minimum value map, the neighborhood direction corresponding to the minimum value is selected to obtain a minimum value direction map.

[0020] In one embodiment, generating a difference map according to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested includes:

[0021] According to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested, the sum of the grayscale value differences between the center point of the pixel and the neighborhood position in each neighborhood direction is calculated;

[0022] The sum of the differences between the grayscale values ​​of the center point and the neighborhood position of the pixel in each neighborhood direction is accumulated to obtain the value of the corresponding pixel, and a difference map is generated based on the value of each pixel.

[0023] In one of the embodiments, the grayscale binarization processing is performed on the difference image according to a preset threshold image to generate a contour binary image, including: setting the grayscale values ​​of pixels in the interpolation image whose grayscale values ​​are greater than the threshold image to the set grayscale values, and setting the grayscale values ​​of the remaining pixels to 0, to generate a contour binary image.

[0024] In one embodiment, the defect detection on the defect contour point set according to the preset minimum value direction map includes:

[0025] Generate the minimum rectangular box R of the package contour according to the defect contour point setk ;

[0026] Find the minimum rectangular box R in the difference map k The minimum value P of the inner pixel value min ;

[0027] Find the minimum value P in the minimum direction map min The gray value of the pixel at position H min ;

[0028] According to the minimum value P min and the pixel gray value H min , for the minimum rectangular frame R k Conduct defect detection.

[0029] In one embodiment, the minimum value P min and the pixel gray value H min , for the minimum rectangular frame R k Conduct defect inspection, including:

[0030] With the minimum value P min The position is centered on the pixel gray value H min Generate bounding regions for directions;

[0031] If the fluctuation of the local mean value in the circumscribed area is within a preset range, the minimum rectangular frame R is determined. k For defects.

[0032] In one embodiment, if the fluctuation of the local mean value in the circumscribed area is within a preset range, then the minimum rectangular frame R is determined. k Defects include:

[0033] Use the affine transformation matrix to transform the pixel points corresponding to the circumscribed area to the horizontal direction, and calculate the pixel grayscale mean R of the circumscribed area after the transformation mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points;

[0034] According to the grayscale mean R mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points, and calculate the minimum rectangular frame R k The defect calculation value Val;

[0035] If the defect calculation value Val is greater than the preset defect minimum threshold, the minimum rectangular frame R is determined. k For defects.

[0036] In one embodiment, the pixel mean R mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points, and calculate the minimum rectangular frame R k The defect calculation value Val includes:

[0037] The minimum value P after conversion min (0,0) The maximum value is MaxP in the grayscale mean of the minimum value of its adjacent points. mean , the minimum value is MinP mean , if P min (0,0) >R mean , then Val is equal to MaxP mean , if P min (0,0) <R mean , then Val is equal to MinP mean .

[0038] A second aspect of the present application provides a wafer defect detection device, comprising:

[0039] An image acquisition module, used to acquire an image of the wafer to be tested;

[0040] An image processing module is used to generate a difference map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the wafer image to be tested; perform grayscale binarization processing on the difference map according to a preset threshold map to generate a contour binary map; perform boundary contour point extraction according to the contour binary map to obtain a defect contour point set; the threshold map is obtained by generating a minimum value map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the set template map, and then performing processing according to the minimum value map;

[0041] The defect detection module is used to perform defect detection on the defect contour point set according to a preset minimum value direction map; the minimum value direction map is obtained by processing according to the minimum value map.

[0042] A third aspect of the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0043] A fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0044] The above-mentioned wafer defect detection method, device, computer equipment and medium obtain the wafer image to be tested, generate a difference map according to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested, and perform grayscale binarization processing on the difference map according to the preset threshold map to generate a contour binary map. Boundary contour points are extracted according to the contour binary map to obtain a defect contour point set, and defect detection is performed on the defect contour point set according to the preset minimum direction map. After the difference map is generated according to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested, the threshold map and minimum direction map obtained by processing the minimum map generated according to the grayscale value of each pixel and the corresponding neighborhood position in the set template map are combined, and the difference map is processed to generate a defect contour point set for defect detection. Compared with using a single threshold for defect point determination, it can reduce the impact of image quality on detection and improve detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of a wafer defect detection method in one embodiment;

[0046] Figure 2 is a flow chart of a wafer defect detection method in another embodiment;

[0047] Figure 3 A flowchart of generating a minimum value map according to the grayscale values ​​of each pixel and the neighboring position in the corresponding neighboring direction in an embodiment;

[0048] Figure 4 A schematic diagram of dividing the neighborhood of a pixel point in one embodiment;

[0049] Figure 5 A schematic diagram of generating a minimum value graph in one embodiment;

[0050] Figure 6 A flowchart of processing a minimum value map to obtain a threshold value map and a minimum value direction map in an embodiment;

[0051] Figure 7 A flowchart of generating a difference map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the wafer image to be tested in one embodiment;

[0052] Figure 8 is a schematic diagram of the gradient direction in one embodiment;

[0053] Fig. 9 A schematic diagram of dual threshold detection division in one embodiment;

[0054] Fig.10 is a flow chart of performing defect detection on a defect contour point set according to a preset minimum value direction map in one embodiment;

[0055] Fig.11is the minimum rectangular frame R in one embodiment k Schematic diagram of the generation of

[0056] Fig.12 A schematic diagram of generating a circumscribed area in one embodiment;

[0057] Fig.13 A schematic diagram of calculating a defect calculation value Val in one embodiment;

[0058] Fig.14 1 is a schematic diagram of a process of a wafer defect detection method in an embodiment;

[0059] Fig.15 It is a structural block diagram of a wafer defect detection device in one embodiment;

[0060] Fig.16 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0062] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0063] In one embodiment, Figure 1 As shown, a wafer defect detection method is provided, comprising:

[0064] Step S140: Acquire an image of the wafer to be tested. The wafer to be tested may be photographed by an image acquisition device to obtain the image of the wafer to be tested.

[0065] Step S150: Generate a difference map based on the grayscale values ​​of each pixel and the corresponding neighborhood position in the wafer image to be tested. The neighborhood position of the pixel can be defined in advance. For example, the area around the pixel can be divided into multiple neighborhood directions, and the position of the spacing step set from the pixel in each neighborhood direction is used as the neighborhood position of the pixel. After acquiring the wafer image to be tested, traverse each pixel of the wafer image to be tested, combine the grayscale values ​​of each pixel and the corresponding neighborhood position, and generate a difference map.

[0066] Step S160: Binarize the difference image according to the preset threshold image to generate a contour binary image. The threshold image is obtained by processing the minimum image based on the grayscale values ​​of each pixel and the corresponding neighborhood position in the set template image. Specifically, the golden template image can be used as the set template image, and each pixel of the set template image is traversed. The minimum image is generated according to the grayscale values ​​of each pixel and the corresponding neighborhood position, and then the threshold image is obtained by processing based on the minimum image. After obtaining the difference image of the wafer to be tested, the grayscale binarization is performed in combination with the threshold image to generate a contour binary image.

[0067] Step S170: Extract boundary contour points according to the contour binary image to obtain a defect contour point set. After obtaining the contour binary image, the Canny operator can be used to perform edge detection to extract contour points to obtain a defect contour point set. The extracted contour points can be grouped, for example, continuous contour points are grouped as one group to obtain multiple groups of defect contour point sets.

[0068] Step S180: performing defect detection on the defect contour point set according to the preset minimum value direction map.

[0069] Among them, the minimum value direction map is obtained by processing according to the minimum value map. After traversing each pixel of the set template image and generating the minimum value map according to the grayscale value of each pixel and the corresponding neighborhood position, the minimum value map can be further processed to obtain the minimum value direction map. For each defect contour point set of the wafer image to be tested, defect detection is performed in combination with the minimum value direction map.

[0070] Furthermore, if Figure 2 As shown, before step S160, the method may further include steps S110 to S130.

[0071] Step S110: traverse each pixel in the set template image, determine the neighborhood of each pixel according to the preset spacing step, and divide the neighborhood into multiple neighborhood directions. Among them, the value of the spacing step is not unique, and can be set according to actual needs. The spacing step can be set to a number of pixel lengths, or it can be set to a specific length value. In this embodiment, 6 pixel lengths are used as the spacing step. The neighborhood of the pixel can be a circle with the pixel position as the center and the spacing step as the radius, and the neighborhood is divided into multiple different neighborhood directions. The positions of the two intersection pixels of the circle in each neighborhood direction of the pixel are two neighborhood positions. The number of neighborhood directions is also not unique, and the neighborhood of the pixel can be divided into 8, 16 or 32 neighborhood directions.

[0072] Step S120: Generate a minimum value map based on the grayscale values ​​of each pixel and the neighboring position in the corresponding neighborhood direction. Traverse each pixel in the set template map, calculate based on the grayscale values ​​of each pixel and the neighboring position in the corresponding neighborhood direction, and obtain a minimum value map.

[0073] In one embodiment, Figure 3 As shown, step S120 includes step S122 and step S124.

[0074] Step S122: According to the grayscale value of each pixel and the neighboring position in the corresponding neighborhood direction, the absolute value of the grayscale difference maximum value of the corresponding pixel in different neighborhood directions is calculated.

[0075] Step S124: Select the minimum value from the absolute values ​​of the grayscale difference maxima of each pixel in different neighborhood directions as the value of the corresponding pixel, and generate a minimum value map according to the values ​​of each pixel.

[0076] like Figure 4 As shown in the figure, traverse each pixel in the set template image, determine the neighborhood of each pixel according to the spacing step, and divide the neighborhood into 16 neighborhood directions, which are recorded as Dir 1, Dir 2, ..., Dir 16. Taking the pixel P0 position as an example, the value of pixel P0 in the minimum value map is Q0, then the calculation formula of Q0 is

[0077] q i =Max(||Pi0-Pi1|-|Pi0-Pi2||,|Pi0-Pi1|,|Pi0-Pi2|),i=1,2,...,16

[0078] Q0=Min(q i ),i=1,2,...,16

[0079] Among them, Pi0 is the gray value of pixel P0, Pi1 and Pi2 are the gray values ​​of the two neighboring positions of pixel P0 in the neighborhood direction i, and q i is the absolute value of the maximum grayscale difference of the three points of pixel P0 in the neighborhood direction i, such as Figure 5 As shown, Diri represents the neighborhood direction i.

[0080] Step S130: Process the minimum value map to obtain a threshold value map and a minimum value direction map. Figure 6 As shown, step S130 includes step S132 and step S134.

[0081] Step S132: Multiply the value of each pixel in the minimum value map by the preset standard deviation threshold, and then perform Gaussian blur processing, set the grayscale value of the pixel less than the set constant to the set constant, and obtain the threshold map. Among them, the values ​​of the standard deviation threshold Std Thrs and the set constant are not unique, and can be set according to actual needs. In this embodiment, in the minimum value map, the value of each pixel is multiplied by the standard deviation threshold Std Thrs, and then perform Gaussian blur of 3*3 pixels. Finally, the grayscale value of the pixel in the map whose grayscale value is less than the constant 4 is set to 4, and the threshold map is obtained.

[0082] Step S134: From the absolute values ​​of the grayscale differences of each pixel in different neighborhood directions in the minimum value map, select the neighborhood direction corresponding to the minimum value to obtain a minimum value direction map. In the minimum value direction map, select the absolute values ​​q of the grayscale differences of the three points of the pixel P0 in the neighborhood direction i. i The neighborhood direction i when it is the minimum value is selected according to each pixel to obtain the minimum value direction map.

[0083] In one embodiment, Figure 7 As shown, step S150 includes step S152 and step S154.

[0084] Step S152: According to the grayscale values ​​of each pixel and the corresponding neighborhood position in the wafer image to be tested, the sum of the grayscale value differences between the center point of the pixel and the neighborhood position in each neighborhood direction is calculated.

[0085] Step S154: Accumulate the sum of the difference between the grayscale value of the center point of the pixel and the neighborhood position in each neighborhood direction to obtain the value of the corresponding pixel, and generate a difference map according to the value of each pixel.

[0086] Specifically, traverse each pixel on the wafer image to be tested, and the value of pixel P0 in the difference map is R0, then the calculation formula of R0 is

[0087] r i =(Pi0-Pi1)+(Pi0-Pi2),i=1,2,...,16

[0088]

[0089] Among them, Pi0 is the gray value of pixel P0, Pi1 and Pi2 are the gray values ​​of the two neighboring positions of pixel P0 in the neighborhood direction i, and r i It is the sum of the differences in grayscale values ​​between the center point of pixel P0 and the neighborhood position in neighborhood direction i.

[0090] Further, step S160 may include: setting the grayscale values ​​of pixels in the interpolation image whose grayscale values ​​are greater than the threshold image to the set grayscale value, and setting the grayscale values ​​of the remaining pixels to 0, to generate a contour binary image. The set grayscale value is not unique and can be set according to actual needs. In this embodiment, the set grayscale value is 255. In the difference image, the grayscale values ​​of some pixels that are greater than the threshold image are set to 255, and the remaining grayscale values ​​are set to 0, to generate a contour binary image.

[0091] After determining the contour binary image, the method of extracting the boundary contour points in step S170 is not unique. Specifically, first, the contour binary image is subjected to Canny boundary contour point extraction. The Canny edge detection algorithm can be divided into five steps: Gaussian filtering, Sobel gradient calculation, non-maximum suppression, double threshold detection, and edge isolated point suppression. The Gaussian function formula is as follows:

[0092]

[0093] Among them, σ represents the Gaussian kernel coefficient, G σ Represents the weight of the non-anchor point coordinate position (x, y) in the Gaussian kernel. Using the Gaussian kernel for image filtering can smooth the image to eliminate noise. The formulas of the Sobel operator's X-direction and Y-direction convolution templates Sx and Sy, and the neighborhood point labeling matrix K of the point to be processed are as follows:

[0094]

[0095] The magnitude of the gradient of each pixel in the image can be calculated by formula (2), as shown in formula (3) to formula (5).

[0096]

[0097] s x =(a2+2a3+a4)-(a0+2a7+a6) (4)

[0098] s y =(a0+2a1+a2)-(a6+2a5+a4) (5)

[0099] like Figure 8 The figure shows the gradient direction diagram, where and The angle θ is the gradient phase angle of point P. Non-maximum suppression is used to compare the gradient modulus of each pixel with the two adjacent pixels in the gradient direction, and the pixel point with the maximum gradient is retained.

[0100] like Fig. 9The figure shows a schematic diagram of dual threshold detection division. Using dual threshold detection, the edge extracted by the high threshold is taken as the main edge, and the low threshold edge connected to the main edge is supplemented in the edge defect part to determine the real and potential edges and form complete edge information. Finally, the edge point detection is completed by suppressing isolated weak edges. According to the gradient direction of the boundary points and the mutual distance, the edge points with the same gradient direction are screened to find the points on the same straight line.

[0101] In one embodiment, Fig.10 As shown, step S180 includes steps S182 to S188.

[0102] Step S182: Generate the minimum rectangular box R of the defect contour according to the defect contour point set k Specifically, after determining the defect contour point set, traverse each defect contour point set, connect each defect contour point set through boundary points, and generate a bounding Rect (the minimum rectangular box covering the contour) as the minimum rectangular box. Assuming that a total of m defect contour point sets are found, the minimum rectangular box of the kth defect contour point set can be represented by the symbol R k .

[0103] Step S184: Find the minimum rectangular box R in the difference map k The minimum value P of the inner pixel value min In the difference map, find the smallest rectangular box R k The minimum value of the inner pixel value is denoted by P min .

[0104] Step S186: Find the minimum value P in the minimum value direction map min The gray value of the pixel at position H min In the minimum direction map, find the minimum value P min The gray value of the pixel at position is denoted as H min Among them, each defect contour point set is regarded as a suspected defect area, and the minimum value P min Indicates the point with the largest jump in each direction in the suspected defect area, the pixel gray value H min Indicates the direction number of the point with the largest jump in the set template graph where the neighborhood changes the least.

[0105] Step S188: According to the minimum value P min and pixel gray value H min , for the minimum rectangular box R k Defect detection is performed. Specifically, the minimum value P min The position is the center, with the pixel gray value H min Generate an external region for the direction; if the fluctuation of the local mean in the external region is within the preset range, the minimum rectangular box R is determinedk The specific value of the preset range can be set according to actual needs, such as Fig.11 As shown in the difference diagram, the minimum value P min The position is the center, the pixel gray value H min For the direction, find a 3*17 circumscribed area, such as Fig.12 The red box area is shown in the figure. Statistical analysis is performed in the outer area to determine whether the fluctuation of the local mean is within the allowable range. If so, the minimum rectangular box R is determined. k It is a defect, otherwise it will be discarded.

[0106] If the fluctuation of the local mean value in the external area is within the preset range, the minimum rectangular box R is determined. k The defect includes: using the affine transformation matrix to transform the pixel points corresponding to the external area to the horizontal direction, and calculating the pixel grayscale mean R of the external area after the transformation mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points; according to the image grayscale mean R mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points, and calculate the minimum rectangular box R k If the defect calculation value Val is greater than the preset defect minimum threshold, the minimum rectangular box R is determined. k For defects.

[0107] Specifically, the affine transformation matrix is ​​used to transform the corresponding pixels of the 3*17 external area to the horizontal direction, and the pixel mean R of the entire external area after the transformation is calculated. mean , standard deviation R std , and the minimum value P after conversion min (0,0) and the minimum value P of its three adjacent points min (-1，-1) , P min (0，-1) , P min (-1，0) .like Fig.13 As shown, the minimum value P after conversion is min (0,0) , P min (-1，-1) , P min (0，-1) , P min (-1，0) is the upper left corner fixed point, and the grayscale mean of the 2*2 area is P mean (0,0) , P mean (-1，-1) , Pmean (0，-1) , P mean (-1，0) . The minimum value P after conversion min (0,0) The grayscale mean of the minimum value of its adjacent points is sorted, and the maximum value is counted as MaxP mean , the minimum value is MinP mean , use Val to represent the minimum rectangular box R k The defect calculation value of min (0,0) >R mean , then Val is equal to MaxP mean , if P min (0,0) <R mean , then Val is equal to MinP mean . Determine the relationship between Val and the minimum threshold of the defect. If it is greater than the minimum threshold, determine that the minimum rectangular box R k It is a defect, otherwise it will be discarded.

[0108] like Fig.14 As shown, in the wafer defect detection method provided by the present application, the image input in the Teach stage is the golden template image, and the image input in the Inspect stage is the wafer image to be inspected. In the Teach stage, traverse each pixel on the golden template image, generate a minimum map according to the grayscale values ​​of each pixel and the corresponding neighborhood position, and further process to obtain a threshold map and a minimum direction map. In the Inspect stage, traverse each pixel on the wafer image to be inspected, generate a difference map according to the grayscale values ​​of each pixel and the corresponding neighborhood position, obtain a contour binary map based on the difference map and the threshold map, and then perform contour extraction on the contour binary map to obtain a defect contour point set, and perform defect detection on the defect contour point set according to the minimum direction map, so that even weak defects can be accurately detected.

[0109] The above wafer defect detection method calculates the minimum value and minimum direction for each pixel in the wafer image to be tested. Each pixel will adaptively generate a threshold range based on anisotropy according to the distribution of its surrounding features. In actual wafer defect detection, the influence of image quality can be overcome to the greatest extent and the detection rate can be improved. The minimum value and difference are calculated by using the neighborhood three-point difference method, which can extract representative features and enhance the features of weak and small defects in the original image.

[0110] This method uses a single pixel as the minimum unit for the minimum value map, minimum value direction map, and difference map. Each pixel can adaptively generate a threshold based on the brightness change difference in the surrounding 16 directions, which can suppress over-inspection problems caused by uneven brightness or sensor noise. In defect determination, the minimum difference point is used to set the minimum value direction in the template map, and statistical analysis of the difference in this direction can improve the robustness of defect detection and reduce the occurrence of over-inspection.

[0111] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0112] Based on the same inventive concept, the embodiment of the present application also provides a wafer defect detection device for implementing the wafer defect detection method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more wafer defect detection device embodiments provided below can refer to the limitations of the wafer defect detection method above, and will not be repeated here.

[0113] In one embodiment, Fig.15 As shown, a wafer defect detection device is provided, comprising: an image acquisition module 110, an image processing module 120 and a defect detection module 130, wherein:

[0114] The image acquisition module 110 is used to acquire the image of the wafer to be tested.

[0115] The image processing module 120 is used to generate a difference map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the wafer image to be tested; perform grayscale binarization processing on the difference map according to a preset threshold map to generate a contour binary map; perform boundary contour point extraction according to the contour binary map to obtain a defect contour point set; the threshold map is obtained by generating a minimum value map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the set template map, and then processing according to the minimum value map.

[0116] The defect detection module 130 is used to perform defect detection on the defect contour point set according to a preset minimum value direction map; the minimum value direction map is obtained by processing according to the minimum value map.

[0117] In one embodiment, the image processing module 120 is also used to traverse each pixel in the set template image, determine the neighborhood of each pixel according to a preset spacing step, and divide the neighborhood into multiple neighborhood directions; generate a minimum value map according to the grayscale value of each pixel and the neighborhood position in the corresponding neighborhood direction; process the minimum value map to obtain a threshold map and a minimum value direction map.

[0118] Each module in the above-mentioned wafer defect detection device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0119] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Fig.16 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a wafer defect detection method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0120] Those skilled in the art will understand that Fig.16 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0121] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps of the above method when executing the computer program.

[0122] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above method are implemented.

[0123] In one embodiment, a computer program product is provided, comprising a computer program, which implements the steps of the above method when executed by a processor.

[0124] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present 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, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0125] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0126] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A wafer defect detection method, characterized in that: include: Acquire an image of a wafer to be tested; Generate a difference map according to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested; The difference image is subjected to grayscale binarization processing according to a preset threshold image to generate a contour binary image; the threshold image is obtained by generating a minimum value image according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the set template image, and then processing according to the minimum value image; Extract boundary contour points according to the contour binary image to obtain a defect contour point set; Defect detection is performed on the defect contour point set according to a preset minimum value direction map; the minimum value direction map is obtained by processing according to the minimum value map.

2. The method according to claim 1, characterized in that Before the grayscale binarization processing is performed on the difference image according to the preset threshold image to generate the contour binary image, the method further includes: Traversing each pixel in the set template image, determining the neighborhood of each pixel according to a preset spacing step, and dividing the neighborhood into multiple neighborhood directions; Generate a minimum value map according to the grayscale values ​​of the pixels and the neighborhood positions in the corresponding neighborhood directions; The minimum value map is processed to obtain a threshold value map and a minimum value direction map.

3. The method according to claim 2, characterized in that The step of generating a minimum value map according to the grayscale values ​​of the pixels and the neighboring positions in the corresponding neighboring directions comprises: According to the grayscale values ​​of the pixels and the neighboring positions in the corresponding neighboring directions, the absolute values ​​of the grayscale differences of the corresponding pixels in different neighboring directions are calculated; From the absolute values ​​of the maximum grayscale differences of each pixel in different neighborhood directions, the minimum value is selected as the value of the corresponding pixel, and a minimum value map is generated according to the values ​​of each pixel.

4. The method according to claim 3, characterized in that: The processing of the minimum value map to obtain a threshold value map and a minimum value direction map includes: Multiply the value of each pixel in the minimum value map by a preset standard deviation threshold, and then perform Gaussian blur processing, and set the grayscale value of pixels less than a set constant to the set constant, to obtain a threshold map; From the absolute values ​​of the maximum grayscale differences of each pixel in different neighborhood directions in the minimum value map, the neighborhood direction corresponding to the minimum value is selected to obtain a minimum value direction map.

5. The method according to any one of claims 1 to 4, characterized in that: The step of generating a difference map according to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested comprises: According to the grayscale value of each pixel and the corresponding neighborhood position in the wafer image to be tested, the sum of the grayscale value differences between the center point of the pixel and the neighborhood position in each neighborhood direction is calculated; The sum of the differences between the grayscale values ​​of the center point and the neighborhood position of the pixel in each neighborhood direction is accumulated to obtain the value of the corresponding pixel, and a difference map is generated based on the value of each pixel.

6. The method according to any one of claims 1 to 4, characterized in that: The grayscale binarization processing is performed on the difference image according to a preset threshold image to generate a contour binary image, including: setting the grayscale values ​​of pixels in the interpolation image whose grayscale values ​​are greater than the threshold image to the set grayscale value, and setting the grayscale values ​​of other pixels to 0, to generate a contour binary image.

7. The method according to any one of claims 1 to 4, characterized in that: The step of performing defect detection on the defect contour point set according to a preset minimum value direction map includes: Generate the minimum rectangular box R of the package contour according to the defect contour point set k ; Find the minimum rectangular box R in the difference map k The minimum value P of the inner pixel value min ; Find the minimum value P in the minimum direction map min The gray value of the pixel at position H min ; According to the minimum value P min and the pixel gray value H min , for the minimum rectangular frame R k Conduct defect detection.

8. The method according to claim 7, characterized in that According to the minimum value P min and the pixel gray value H min , for the minimum rectangular frame R k Conduct defect detection, including: With the minimum value P min The position is centered on the pixel gray value H min Generate bounding regions for directions; If the fluctuation of the local mean value in the circumscribed area is within a preset range, the minimum rectangular frame R is determined. k For defects.

9. The method according to claim 8, characterized in that If the fluctuation of the local mean value in the circumscribed area is within a preset range, the minimum rectangular frame R is determined. k Defects include: Use the affine transformation matrix to transform the pixel points corresponding to the circumscribed area to the horizontal direction, and calculate the pixel grayscale mean R of the circumscribed area after the transformation mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points; According to the grayscale mean R mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points, and calculate the minimum rectangular frame R k The defect calculation value Val; If the defect calculation value Val is greater than the preset defect minimum threshold, the minimum rectangular frame R is determined. k For defects.

10. The method according to claim 9, characterized in that According to the pixel mean R mean , the minimum value after conversion P min (0,0) and the minimum value of its adjacent points, and calculate the minimum rectangular frame R k The defect calculation value Val includes: The minimum value P after conversion min (0,0) The maximum value is MaxP in the grayscale mean of the minimum value of its adjacent points. mean , the minimum value is MinP mean , if P min (0,0) >R mean , then Val is equal to MaxP mean , if P min (0,0) <R mean , then Val is equal to MinP mean .

11. A wafer defect detection device, characterized in that: include: An image acquisition module, used to acquire an image of the wafer to be tested; An image processing module is used to generate a difference map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the wafer image to be tested; perform grayscale binarization processing on the difference map according to a preset threshold map to generate a contour binary map; perform boundary contour point extraction according to the contour binary map to obtain a defect contour point set; the threshold map is obtained by generating a minimum map according to the grayscale values ​​of each pixel and the corresponding neighborhood position in the set template map, and then performing processing according to the minimum map; The defect detection module is used to perform defect detection on the defect contour point set according to a preset minimum value direction map; the minimum value direction map is obtained by processing according to the minimum value map.

12. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 10 are implemented.

13. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 10 are implemented.

Images