An image recognition method, device and system for wafer bonding

Abstract

The application relates to an image recognition method, device and system for wafer bonding. The method according to some embodiments of the application comprises: providing a template image; detecting edges in the template image; screening the detected edges to obtain an outer edge of a template region in the template image; and measuring the position of the template region in the template image. Compared with the prior art, the application can automatically screen the template region in the template image, and can also obtain highly consistent templates for multiple cameras, thereby improving the image recognition accuracy.

Description

Technical Field

[0001] This application generally relates to the field of semiconductor technology, and more specifically, to an image recognition method, apparatus, and system for wafer bonding. Background Technology

[0002] In the manufacturing or process inspection of large-scale integrated circuits in semiconductors, semiconductor wafers undergo multiple manufacturing stages and are processed or inspected by various different devices. Wafer bonding technology has become a key technology for the integrated development and practical application of semiconductor manufacturing technology. Wafer bonding refers to the process of bonding two flat wafers face to face and applying certain external conditions such as pressure, temperature, and voltage to generate atomic or molecular bonding forces at the interface between the two wafers, such as covalent bonds, metallic bonds, and molecular bonds. This allows the bonding strength between the two surfaces to reach a certain level, thus making the two wafers a single unit.

[0003] In wafer bonding processes, two wafers need to be aligned and bonded. This is typically achieved using a template image. However, existing template image selection usually requires the user to manually select the template area, and the position and size of the template area need to be manually set and adjusted. The adjustment process requires human judgment, which leads to problems such as low accuracy and poor repeatability of the recognition results.

[0004] Therefore, it is necessary to improve the existing solutions in order to solve the problems existing in the existing technology. Summary of the Invention

[0005] One of the objectives of this application is to provide an image recognition method, apparatus, and system for wafer bonding, which solves the problems of low recognition accuracy and poor repeatability caused by users manually selecting template areas and by rotating and / or blurring template images.

[0006] According to one aspect of this application, an image recognition method for wafer bonding is provided, comprising: providing a template image; detecting edges in the template image; filtering the detected edges to obtain the outer edge of a template region in the template image; and measuring the position of the template region in the template image. According to another aspect of this application, an image recognition apparatus for wafer bonding is provided, comprising: a supply module configured to provide a template image; a detection module configured to detect edges in the template image; a filtering module configured to filter the detected edges to obtain the outer edge of a template region in the template image; and a measurement module configured to measure the position of the template region in the template image.

[0007] According to one aspect of this application, this application also provides a non-volatile computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the image recognition method for wafer bonding described in this application.

[0008] According to one aspect of this application, an image recognition system for wafer bonding is also provided, comprising: a processor; a non-volatile computer-readable storage medium storing computer-executable instructions coupled to the processor; and a stage for supporting a wafer; wherein the processor is configured to execute the computer-executable instructions to implement the image recognition method for wafer bonding described in this application on the wafer. The image recognition method, apparatus, and system for wafer bonding provided in this application can automatically filter out template regions in a template image, improving image recognition accuracy. Furthermore, the image recognition method, apparatus, and system for wafer bonding provided in this application can obtain highly consistent templates for multiple cameras. Additionally, the image recognition method, apparatus, and system for wafer bonding provided in this application can more accurately calculate the center position and boundaries of the template image, making the template less susceptible to damage during subsequent image processing, further improving image recognition accuracy. Attached Figure Description

[0009] The accompanying drawings, necessary for describing embodiments of this application or the prior art, will be briefly described below to facilitate the depiction of embodiments of this application. It is obvious that the drawings described below represent only a portion of the embodiments in this application. Those skilled in the art will be able to derive other embodiments from the illustrations in these drawings without requiring inventive effort.

[0010] Figure 1 This is a flowchart of an image recognition method for wafer bonding according to some embodiments of this application.

[0011] Figure 2 This is a schematic diagram of a grayscale template image according to some embodiments of this application.

[0012] Figure 3 This is a schematic diagram of a template image after performing an edge detection step according to some embodiments of this application.

[0013] Figure 4 This is a flowchart illustrating edge processing according to some embodiments of this application.

[0014] Figure 5 For implementation according to some embodiments of this application Figure 4 A schematic diagram of the template image after the edge processing steps shown.

[0015] Figure 6 This is a flowchart illustrating the specific location of the measurement template area according to some embodiments of this application.

[0016] Figure 7 For implementation according to some embodiments of this application Figure 6 A schematic diagram of the template image after the step of measuring the position of the template area is shown.

[0017] Figure 8 This is a structural block diagram of an image recognition device for wafer bonding according to some embodiments of this application. Detailed Implementation

[0018] To better understand the spirit of this application, the following description is based on some preferred embodiments of this application.

[0019] Various embodiments of this application are discussed in detail below. Although specific embodiments are discussed, it should be understood that these embodiments are for illustrative purposes only. Those skilled in the art will recognize that other components and configurations can be used without departing from the spirit and scope of this application.

[0020] Figure 1 This is a flowchart of an image recognition method for wafer bonding according to some embodiments of this application. The following is in conjunction with... Figures 2 to 7 To describe in detail, such as Figure 1 The image recognition method for wafer bonding shown in some embodiments of this application is illustrated.

[0021] First, the image is read in (step S10). Specifically, a template image is selected and read in.

[0022] In some embodiments, the image is grayscaled (step S20). In this step, the read template image is grayscaled to obtain a grayscaled template image. Figure 2 This is a schematic diagram of a grayscale template image according to some embodiments of this application. Figure 2 As shown, template image 20 includes template region 22.

[0023] In some embodiments, image contrast is enhanced (step S30). In some embodiments, an operation is performed on the grayscaled template image to improve the contrast of the template image.

[0024] In some embodiments, the image is sharpened (step S40). In some embodiments, the template image processed in step S30 is sharpened.

[0025] Steps S30 and S40 can be used to improve the accuracy of edge detection. In some embodiments, steps S30 and S40 can be omitted, and edge detection can be performed solely based on the grayscale template image.

[0026] In some embodiments, edge detection is performed (step S50). In some embodiments, the template image from step S40 is processed to detect and locate all edges in the template image. In some embodiments, an edge detection algorithm based on the Sobel operator is used to detect and locate all edges in the template image, specifically as follows:

[0027] Assuming the template image is a two-dimensional structure, the definition of the partial derivative in the x-direction is: The same applies in the y-direction;

[0028] Now, assuming Δx = 1, the above definition can be simplified to:

[0029] Furthermore, formula (1) can be rewritten as:

[0030]

[0031]

[0032] Combining the partial derivatives in the x-direction and the partial derivatives in the y-direction, we can obtain the gradient G:

[0033]

[0034] Applying the Sobel operator, the weights are 1 in the [1,1] and [-1,1] directions, and 4 in the [0,1] and [1,0] directions. Therefore, the calculation formulas in the x and y directions are:

[0035]

[0036]

[0037] Based on the above calculation formulas (5) and (6), they are converted into 3*3 convolution kernels as follows:

[0038] Table 1:

[0039] -1 0 +1 -2 0 +2 -1 0 +1

[0040] Sobel weights in the x-direction

[0041] Table 2:

[0042] -1 -2 -1 0 0 0 +1 +2 +1

[0043] Sobel weights in the x-direction

[0044] Table 3:

[0045] (x-1, y-1) (x, y-1) (x+1, y-1) (x-1,y) (x,y) (x+1,y) (x-1, y+1) (x,y+1) (x,y+1)

[0046] The pixel at (x,y) and its neighborhood in the template image

[0047] To determine whether a pixel in the template image is an edge pixel: calculate the magnitude of the gradient G using the formulas (4), (5) and (6) above; then, set a threshold T according to the user's needs; when G is greater than the threshold T, the current pixel is considered to be an edge pixel; when G is less than the threshold T, the current pixel is not an edge pixel.

[0048] Figure 3 This is a schematic diagram of a template image after performing an edge detection step according to some embodiments of this application. Figure 3 As shown, template image 30 includes template region 32. After performing the edge detection step S50, multiple edges are detected in the template image. Specifically, as... Figure 3 As shown, it includes the outer edge 322 and inner edge 324 of the template area 32, the edge 326 located outside the outer edge 322 of the template area 32, and the edge 328.

[0049] In some embodiments, the edges are processed (step S60). See also Figure 1 and Figure 3 In step S60, all detected edges are filtered to finally obtain the outer edge 322 of the template region 32 in the template image 30. Figure 4 This is a detailed flowchart of edge processing according to some embodiments of this application.

[0050] In some embodiments, edges in the template image with a length less than a threshold are removed (step S602). The main purpose of this step is to remove small edges caused by noise, such as... Figure 3 Edge 328 in the image. Noise-generated edges are generally very short, and the calculated gradient G is very large, making it impossible to filter them out by adjusting the threshold T. Therefore, in some embodiments of this application, an edge length filtering method is used to remove these noisy edges 328. For example, in some embodiments of this application, when the detected edge length threshold is less than 10 pixels, the edge is removed. However, it should be understood that 10 pixels here is merely a preferred embodiment of this application; in other cases, the edge length threshold can be selected as needed, and no specific limitation is made here.

[0051] In some embodiments, the broken edges in the template image are repaired (step S604). Due to defects in the edge detection algorithm or factors such as lighting and image blurring, breaks may occur at certain locations on the detected edges of the template image 30. For example, ... Figure 3 As shown, a break 322a exists at the corner of the outer edge 322 of the template region 32. In this case, the broken outer edge 322 needs to be supplemented. Typically, the break is located at a corner. However, in some embodiments, the break may also occur in other smooth areas, such as the middle section of the outer edge 322. In some embodiments, the broken edge in the template image 30 is supplemented as follows: when an edge in the template image 30 is detected to be not a complete edge contour, a scan is performed with one endpoint of the broken edge as the center; when another endpoint of the broken edge is found within the scan area, the two endpoints are connected. In some embodiments, the threshold is set to 10, i.e., the scan radius is 10 pixels. Therefore, the scan area is a circle with a radius of 10 pixels centered on one endpoint of the broken edge. However, it should be understood that 10 pixels here is merely a preferred embodiment of this application; in other cases, the scan radius can be selected as needed, and no specific limitation is made here.

[0052] In some embodiments, unclosed edges in the template image are removed (step S606). Due to factors such as lighting, larger noisy edges may also be detected, for example... Figure 3 Edge 326 in the image. Typically, these noisy edges are unclosed. Therefore, edge 326 can be removed by determining whether the detected edges are closed. When an unclosed edge is detected, it is removed.

[0053] In some embodiments, the inner edge contour of the edge contour formed by the closed edges in the template image is removed (step S608). Specifically, the position coordinates of the outer edge 322 of the template region in the template image are obtained to obtain the closed outer edge contour enclosed by the outer edge 322; then the edges (i.e., inner edges 324) within the contour region enclosed by the outer edge 322 are removed.

[0054] Figure 5 For implementation according to some embodiments of this application Figure 4 A schematic diagram of the template image after the edge processing steps shown. See also... Figure 3 and Figure 5 After the edge processing step, the template image 30 retains only the outer edge 322 of the template region 32. Noisy edges 328 and unclosed edges 326 are removed, and the break 322a at the corner of the outer edge 322 of the template region 32 is repaired and supplemented.

[0055] See Figure 1 and Figure 5 In some embodiments, the position of the template region is measured (step S70). In some embodiments, the position of the template region 32 in the template image 30 is measured based on the distance between the outer edge 332 of the template region 32 and the boundary of the template image 30. Figure 6 This is a flowchart illustrating the specific location of the measurement template area according to some embodiments of this application.

[0056] In some embodiments, the pixel coordinates of the outer edge of the template region are obtained (step S702). In some embodiments, the outer edge 322 of the template region 32 obtained in step S60 is extracted, and the pixel coordinates at the following four positions are found by comparing all extracted outer edges 322: (1) the x-coordinate of the pixel with the minimum x-coordinate; (2) the y-coordinate of the pixel with the minimum y-coordinate; (3) the x-coordinate of the pixel with the maximum x-coordinate; and (4) the y-coordinate of the pixel with the maximum y-coordinate. In some embodiments, the origin of the coordinate axis is located at the upper left corner of the template image 30.

[0057] In some embodiments, the distance between the outer edge of the template region and the boundary of the template image is calculated (step S704). In some embodiments, the distances at the following four locations are calculated: (1) the distance between the x-value of the minimum pixel and the left boundary of the template image 30; (2) the distance between the y-value of the minimum pixel and the upper boundary of the template image 30; (3) the distance between the x-value of the maximum pixel and the right boundary of the template image 30; and (4) the distance between the y-value of the maximum pixel and the lower boundary of the template image 30.

[0058] Through the above steps S702 and S704, the position of the template region 32 in the template image 30 can be measured and determined.

[0059] Figure 7 For implementation according to some embodiments of this application Figure 6 This is a schematic diagram of the template image after the step of measuring the location of the template area. (See diagram below.) Figure 7 As shown, the distance between the outer edge 722 of the template region 72 and the boundary of its corresponding template image 70 are not the same, which means that the center of the template region 72 is not located at the center of the template image 70.

[0060] In some embodiments, when the center of the template region 72 is not located at the center of the template image 70, the center of the template region 72 is located at the center of the template image 70. For example, the center of the template region 72 can be moved to the center of the template image 70 by moving it, or the boundary of the template image 70 can be cropped so that the distance between the outer edge 722 of the template region 72 and the cropped boundary of the template image 70 is the same.

[0061] See back Figure 1 In some embodiments, the template region position and parameters are marked (step S80). In some embodiments, the template region is marked when its center is located at the center of the template image. Specifically, if the distance between the x-position value of the minimum pixel and the left boundary of the template image is equal to the distance between the x-position value of the maximum pixel and the right boundary of the template image, and the distance between the y-position value of the minimum pixel and the upper boundary of the template image is equal to the distance between the y-position value of the maximum pixel and the lower boundary of the template image, then the center of the template region is located at the center of the template image. That is, see [link to documentation]. Figure 5 When the distance between the pixel with the minimum X-coordinate in the outer edge 322 of template region 32 and the left boundary of template image 30 is equal to the distance between the pixel with the maximum X-coordinate in the outer edge 322 of template region 32 and the right boundary of template image 30, and when the distance between the pixel with the minimum Y-coordinate in the outer edge 322 of template region 32 and the upper boundary of template image 30 is equal to the distance between the pixel with the maximum Y-coordinate in the outer edge 322 of template region 32 and the lower boundary of template image 30, it indicates that the center of template region 32 is located at the center of template image 30. After detecting that the center of the template region is located at the center of the template image, the template region is located and identified. Specifically, after detecting that the center of the template region is located at the center of the template image, the position and corresponding parameters of the template region are identified.

[0062] Figure 8 This is a structural block diagram of an image recognition apparatus for wafer bonding according to some embodiments of this application. Figure 8As shown, the image recognition device 80 for wafer bonding includes a supply module 802, a detection module 804, a screening module 806, a measurement module 808, an identification module 810, a preprocessing module 812, and an adjustment module 814. The image recognition method for wafer bonding according to this application embodiment can be implemented by the image recognition device 80 for wafer bonding according to this application embodiment. Specifically, steps S10 and S20 can be implemented by the supply module 802; steps S30 and S40 can be implemented by the preprocessing module 812; step S50 can be implemented by the detection module 804; step S60 can be implemented by the screening module 806; step S70 can be implemented by the measurement module 808; and step S80 can be implemented by the identification module 810. Furthermore, when the center of the template region is not located at the center of the template image, the adjustment module 814 also ensures that the center of the template region is located at the center of the template image. In other words, the image recognition device 80 for wafer bonding can implement any step of any method described herein via the supply module 802, detection module 804, screening module 806, measurement module 808, marking module 810, preprocessing module 812, and adjustment module 814, including Figure 1 , Figure 4 as well as Figure 6 The method steps are described below.

[0063] Other embodiments of this application relate to a non-volatile computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement any step of any method described herein, including Figure 1 , Figure 4 as well as Figure 6 The method steps described herein. The program can be implemented in any of a variety of ways, including program-based technologies, component-based technologies, and / or object-oriented technologies. For example, the program may be implemented using ActiveX controls, C++ objects, JavaBeans, Microsoft Foundation Classes (MFC), Streaming SIMD Extensions (SSE), or other technologies or methods as needed.

[0064] Additionally, some embodiments of this application provide an image recognition system for wafer bonding. The system includes a processor, a non-volatile computer-readable storage medium storing computer-executable instructions, and a stage. The non-volatile computer-readable storage medium storing the computer-executable instructions is coupled to the processor. The stage can be used to support a wafer. The processor is configured to execute the computer-executable instructions to implement any step of any of the methods described herein on the wafer, including... Figure 1 , Figure 4 as well as Figure 6 The method steps are described below.

[0065] The image recognition method, apparatus and system for wafer bonding provided in this application have at least the following advantages: (1) the specific location of the template region in the template image can be found through the edge detection algorithm; (2) the edge outside the template region can be found through the edge detection algorithm; (3) broken edges can be supplemented; (4) the location of the template region in the template image can be determined by measuring the distance between the template region and the boundary of the template image; (5) the template outline can be found by the edge detection algorithm, and the template region can be automatically cropped; (6) the accuracy of edge detection can be improved by improving the contrast and edge sharpening of the template image; (7) by determining the location of the template image, a method for multiple cameras to obtain a highly consistent template image is provided.

[0066] It should be noted that throughout this specification, the reference to "some embodiments of this application" or similar terms means that a particular feature, structure, or characteristic described together with other embodiments is included in at least one embodiment and may not necessarily be presented in all embodiments. Therefore, the corresponding appearance of the phrase "some embodiments of this application" or similar terms throughout this specification does not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics of any particular embodiment may be combined with one or more other embodiments in any suitable manner.

[0067] The technical content and features of this invention have been disclosed above. However, those skilled in the art may still make various substitutions and modifications based on the teachings and disclosures of this invention without departing from the spirit of this invention. Therefore, the scope of protection of this invention should not be limited to the content disclosed in the embodiments, but should include various substitutions and modifications that do not depart from this invention, and should be covered by the claims of this patent application.

Claims

1. An image recognition method for wafer bonding, characterized in that, The image recognition method for wafer bonding includes: Provide template images; Detect edges in the template image; The detected edges are filtered to obtain the outer edges of the template region in the template image. Specifically, when an edge in the template image is detected as not a complete edge contour, a scan is performed with one endpoint of the broken edge as the center. When another endpoint of the broken edge is found within the scanned area, the two endpoints are connected. The position of the template region in the template image is measured based on the distance between the outer edge of the template region and the boundary of the template image.

2. The image recognition method for wafer bonding according to claim 1, characterized in that, The template region is identified when its center is located at the center of the template image.

3. The image recognition method for wafer bonding according to claim 2, characterized in that, When the distance between the pixel with the minimum horizontal coordinate in the outer edge of the template region and the left boundary of the template image is equal to the distance between the pixel with the maximum horizontal coordinate in the outer edge of the template region and the right boundary of the template image, and when the distance between the pixel with the minimum vertical coordinate in the outer edge of the template region and the upper boundary of the template image is equal to the distance between the pixel with the maximum vertical coordinate in the outer edge of the template region and the lower boundary of the template image, the center of the template region is located at the center of the template image.

4. The image recognition method for wafer bonding according to claim 1, characterized in that, Before detecting edges in the template image, the method further includes: Increase the contrast of the template image; and Sharpen the template image.

5. The image recognition method for wafer bonding according to claim 1, characterized in that, Edge detection algorithms are used to detect edges in the template image.

6. The image recognition method for wafer bonding according to claim 5, characterized in that, The edge detection algorithm is based on the Sobel operator.

7. The image recognition method for wafer bonding according to claim 5, characterized in that, It also includes at least one of the following steps: Remove edges from the template image whose length is less than a threshold; Remove unclosed edges from the template image; and Remove the inner edge contours from the edge contours formed by the closed edges in the template image.

8. The image recognition method for wafer bonding according to claim 7, characterized in that, The threshold is a length of 10 pixels.

9. The image recognition method for wafer bonding according to claim 1, characterized in that, The radius of the scanning area is 10 pixels in length.

10. The image recognition method for wafer bonding according to claim 7, characterized in that, The inner edge contours of the edge contours formed by closed edges in the template image are removed using the following method: Obtain the position coordinates of the outer edge of the template region in the template image; and Remove the edges within the contour area enclosed by the outer edge.

11. The image recognition method for wafer bonding according to claim 1, characterized in that, When the center of the template region is not located at the center of the template image, the center of the template region is located at the center of the template image.

12. An image recognition device for wafer bonding, characterized in that, The image recognition device for wafer bonding includes: A supply module configured to provide template images; A detection module configured to detect edges in the template image; A filtering module is configured to: filter the detected edges to obtain the outer edges of the template region in the template image, wherein when an edge in the template image is detected to be a non-complete edge contour, a scan is performed with one endpoint of the broken edge as the center; and when another endpoint of a broken edge is found within the scanned area, the two endpoints are connected; and A measurement module configured to measure the position of the template region in the template image based on the distance between the outer edge of the template region and the boundary of the template image.

13. The image recognition device for wafer bonding according to claim 12, characterized in that, The image recognition device for wafer bonding further includes an identification module configured to identify the template region when the center of the template region is located at the center of the template image.

14. The image recognition device for wafer bonding according to claim 13, characterized in that, The identification module is further configured to: When the distance between the pixel with the minimum horizontal coordinate in the outer edge of the template region and the left boundary of the template image is equal to the distance between the pixel with the maximum horizontal coordinate in the outer edge of the template region and the right boundary of the template image, and when the distance between the pixel with the minimum vertical coordinate in the outer edge of the template region and the upper boundary of the template image is equal to the distance between the pixel with the maximum vertical coordinate in the outer edge of the template region and the lower boundary of the template image, the center of the template region is determined to be located at the center of the template image.

15. The image recognition device for wafer bonding according to claim 12, characterized in that, The image recognition apparatus for wafer bonding further includes a preprocessing module configured to: before detecting edges in the template image: Increase the contrast of the template image; and Sharpen the template image.

16. The image recognition device for wafer bonding according to claim 15, characterized in that, The detection module is configured to use an edge detection algorithm to detect edges in the template image.

17. The image recognition device for wafer bonding according to claim 16, characterized in that, The edge detection algorithm is based on the Sobel operator.

18. The image recognition device for wafer bonding according to claim 16, characterized in that, The filtering module is also configured to: Remove edges from the template image whose length is less than a threshold; Remove unclosed edges from the template image; and Remove the inner edge contours from the edge contours formed by the closed edges in the template image.

19. The image recognition device for wafer bonding according to claim 18, characterized in that, The threshold is a length of 10 pixels.

20. The image recognition device for wafer bonding according to claim 12, characterized in that, The radius of the scanning area is 10 pixels in length.

21. The image recognition device for wafer bonding according to claim 18, characterized in that, The filtering module is further configured to remove the internal edge contours from the edge contours formed by closed edges in the template image in the following manner: Obtain the position coordinates of the outer edge of the template region in the template image; and Remove the edges within the contour area enclosed by the outer edge.

22. The image recognition device for wafer bonding according to claim 12, characterized in that, The image recognition device for wafer bonding further includes an adjustment module configured to center the template region at the center of the template image when the center of the template region is not located at the center of the template image.

23. A non-volatile computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores one or more programs that can be executed by one or more processors to implement the image recognition method for wafer bonding as described in any one of claims 1-11.

24. An image recognition system for wafer bonding, comprising: processor; A non-volatile computer-readable storage medium storing computer-executable instructions, coupled to the processor; as well as A stage, used to support a wafer; The processor is configured to execute the computer-executable instructions to implement the image recognition method for wafer bonding according to any one of claims 1-11 on the wafer.

Images