Image-based die automatic positioning method, device, storage medium and program product

Abstract

The application discloses a kind of image-based die automatic positioning method, with the physical position of wafer gap center as origin, a series of images are shot along the vertical axis of rectangular coordinate system, and these images are spliced into complete vertical axis wafer map;Based on the vertical axis wafer map, the edge position of wafer and the size and position of each die are determined;According to the die distribution on the vertical axis, the die distribution data of the whole wafer is further inferred.The present application is efficient, automatic, accurate, and can adapt to different sizes and types of dies with clear edges, effectively overcoming the problems of low efficiency and poor adaptability in the prior art.

Description

Technical Field

[0001] This invention belongs to the field of image processing and defect detection, and specifically relates to automatic grain positioning technology. Background Technology

[0002] In the semiconductor manufacturing industry, automated optical inspection of wafer defects plays a crucial role in ensuring chip yield. This technology primarily focuses on two core steps: accurately determining the location of all dies on the wafer, and performing high-resolution defect identification based on this. Currently, the industry's commonly used technical approaches can be broadly categorized as follows, but each faces significant challenges: (1) Full-field scanning and positioning: This technology performs full-field imaging of the wafer and combines image processing algorithms to locate the die. However, on the one hand, full-field scanning or multiple images require a lot of time, resulting in extremely low detection efficiency, which is difficult to meet the urgent need for rapid detection in large-scale production; on the other hand, complex mechanical motion control increases the complexity and cost of the system, and the accuracy and stability of the mechanical device will also have a significant impact on the detection results, thereby reducing the accuracy of the detection.

[0003] (2) Template matching-based localization: This method determines the position of the die by matching a known die template with an image. However, it is highly dependent on precise prior knowledge such as die size and spacing, and requires pre-collection and saving of templates. In actual production, wafer size and die type vary greatly, and the automation level of pre-acquiring and accurately configuring these parameters is low, resulting in poor system adaptability and inability to flexibly cope with diverse production scenarios.

[0004] (3) Auxiliary positioning device: A physical transparent plate with coordinate marks is used to cover the wafer to assist in coordinate mapping. However, this method has many drawbacks. It is not an automated operation, introduces additional physical steps, is not suitable for online high-speed inspection, and the physical transparent plate may contaminate the wafer. Summary of the Invention

[0005] To address the technical problems mentioned in the background section, this invention proposes an image-based automatic die positioning method, device, storage medium, and program product.

[0006] To achieve the above-mentioned technical objectives, the technical solution of the present invention is as follows: An image-based automatic die positioning method establishes a Cartesian coordinate system with the physical location of the wafer notch center as the origin, takes a series of images along the vertical axis of the Cartesian coordinate system, and stitches these images into a complete vertical axis wafer image; determines the edge position of the wafer and the size and position of each die based on the vertical axis wafer image; and infers the die distribution data of the entire wafer based on the die distribution on the vertical axis.

[0007] Furthermore, the above method includes the following steps: (1) Determine the physical location of the notch center of the wafer to be tested, and establish a rectangular coordinate system in the plane where the wafer is located with the location as the origin. Take a series of images along the vertical axis of the coordinate system and stitch them together to form a complete vertical axis wafer image. (2) Convert the vertical axis wafer image into a grayscale image; (3) Binarize the grayscale image to obtain a binarized image, search the binarized image to obtain the upper edge point set and lower edge point set of the wafer, and take the point with the smallest vertical axis value in the lower edge point set as the position of the wafer notch center on the image; (4) Find straight lines in the grayscale image to obtain a straight line image; perform morphological closing operation on the straight lines in the straight line image, merge adjacent parallel lines to obtain horizontal fusion lines and vertical fusion lines. The width of the fusion lines is the grain gap width w1, and the intersection of the horizontal fusion lines and the vertical fusion lines is the grain corner point. Multi-target matching is performed on the grayscale image based on known grain corner points to obtain a series of grain corner points; (5) Based on the series of grain corner points obtained in step (4), calculate the average value of the horizontal adjacent corner points to obtain the grain width, calculate the average value of the vertical adjacent corner points to obtain the grain height, and calculate the grain deflection angle based on the position of the grain corner points. (6) Based on the data calculated in step (5), the column of grains closest to the position of the wafer notch center obtained in step (3) on the image is taken as the grain column on the vertical axis of the wafer; (7) The grain distribution of the entire wafer can be inferred from the position of the grains on the vertical axis of the wafer.

[0008] Further, in step (3), a sliding window is set to move along the grayscale image, and the mean value within the sliding window is used as a threshold for binarization to obtain the binarized image.

[0009] Furthermore, morphological operations are performed on the binarized image to denoise it.

[0010] Further, in step (4), a grain corner is selected as the center on the grayscale image, and an image of size w2*w2 is used as the template image. Multi-target matching is performed in the grayscale image by the correlation coefficient method to obtain a series of grain corners, where w2>w1.

[0011] Secondly, the present invention also provides a computer device comprising a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the above-described image-based automatic grain positioning method.

[0012] Furthermore, the present invention also provides a computer storage medium storing a computer program, which, when executed on a processor, implements the above-described image-based automatic die positioning method.

[0013] Finally, the present invention also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the above-described image-based automatic die positioning method.

[0014] The beneficial effects of adopting the above technical solution are as follows: Unlike existing technologies such as full-field scanning and positioning, template matching-based positioning, and assisted positioning devices, the technical solution of this invention has significant advantages. It only requires knowledge of the basic parameter of wafer size and only requires capturing a grain image along a single path on the wafer to determine the position of all grains on the wafer. This invention is efficient, automatic, and accurate, and can adapt to grains of different sizes and types with well-defined edges, effectively overcoming the problems of low efficiency and poor adaptability in existing technologies. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method in an embodiment; Figure 2 These are the grayscale images and their corresponding binarized images from the embodiments; Figure 3 This is a wafer top edge image obtained from the embodiment; Figure 4 This is a wafer bottom edge image obtained from the embodiment; Figure 5 This is a line graph obtained from the embodiment; Figure 6 This is a linear fusion diagram obtained from the embodiment; Figure 7 This is a schematic diagram of the grain corners obtained in the example; Figure 8 These are a series of grain diagrams obtained from the examples. Detailed Implementation

[0016] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0017] This embodiment describes an automatic grain positioning method for images, the overall steps of which are as follows: Figure 1 As shown.

[0018] Step 1: Take a series of images along the vertical axis of the wafer and stitch them together to form a complete vertical axis wafer image: Determine the physical location of the notch center of the wafer to be tested. Using this location as the origin, establish a rectangular coordinate system in the plane where the wafer is located. Take a series of images along the vertical axis of the coordinate system and stitch them together to form a complete vertical axis wafer image.

[0019] Step 2: Convert the vertical axis wafer image into a grayscale image G.

[0020] Step 3: Locate the intersection of the vertical axis and the edge of the wafer: (301) A sliding window of a certain size (99 pixels in this embodiment) is moved across the grayscale image G, and binarization is performed using the mean value within the sliding window as a threshold to obtain a binarized image B of the grayscale image G. Then, morphological operations are used to denoise the binarized image B, such as... Figure 2 As shown.

[0021] (302) On the binarized image B, search from top to bottom and record the position of the first non-zero pixel in each column to obtain the set of points on the upper edge of the wafer, as shown below. Figure 3 As shown; searching from bottom to top, recording the position of the first non-zero pixel in each column, obtaining the set of bottom edge points of the wafer. The point with the smallest vertical axis value in this set is taken as the position of the wafer notch center on the image, for later use, as shown. Figure 4 As shown.

[0022] Step 4: Determine the location of the grains: (401) Since the grain edges are roughly rectangular, a straight line is found in the grayscale image G to obtain the straight line image L, as shown below. Figure 5 As shown.

[0023] (402) Perform morphological closing operation on the straight lines in the line graph L, and merge adjacent parallel lines. The width of the merged line is the intergranular gap width w1, as shown below. Figure 6 As shown, the intersection of the horizontal and vertical fusion lines is the corner point of the grain. Figure 7 As shown.

[0024] (403) Select a grain corner point as the center on the grayscale image G, and use an image of size w2*w2 as the template image. Perform multi-target matching in the grayscale image using the correlation coefficient method to obtain a series of grain corner points, where w2>w1, such as Figure 8 As shown.

[0025] Step 5: Calculate the grain size and deflection angle: The average of the horizontally adjacent corner points is calculated to obtain the grain width, the average of the vertically adjacent corner points is calculated to obtain the grain height, and the grain deflection angle is calculated based on the position of the grain corner points.

[0026] Step 6: Determine the grain rows on the vertical axis of the wafer: Based on the data calculated in step 5, the column of grains closest to the position of the wafer notch center obtained in step 3 on the image is taken as the grain column on the vertical axis of the wafer.

[0027] Step 7: Determine the grain distribution on the wafer: Based on the regular arrangement and identical size of the grains on the same wafer, the grain distribution of the entire wafer is calculated using the grain positions along the vertical axis.

[0028] This embodiment also relates to a computer device, which includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the above-described image-based automatic grain positioning method.

[0029] This embodiment also relates to a computer storage medium storing a computer program, which, when executed on a processor, implements the above-described image-based automatic die positioning method.

[0030] This embodiment also relates to a computer program product, including a computer program or instructions, which, when executed by a processor, implement the above-described image-based automatic die positioning method.

[0031] The embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. An image-based automatic grain positioning method, characterized in that: A rectangular coordinate system is established with the physical location of the wafer notch center as the origin. A series of images are taken along the vertical axis of the rectangular coordinate system and these images are stitched together to form a complete vertical axis wafer image. Based on this vertical axis wafer image, the edge position of the wafer and the size and position of each grain are determined. Based on the grain distribution on the vertical axis, the grain distribution data of the entire wafer is inferred.

2. The image-based automatic grain positioning method according to claim 1, characterized in that, Includes the following steps: (1) Determine the physical location of the notch center of the wafer to be tested, and establish a rectangular coordinate system in the plane where the wafer is located with the location as the origin. Take a series of images along the vertical axis of the coordinate system and stitch them together to form a complete vertical axis wafer image. (2) Convert the vertical axis wafer image into a grayscale image; (3) Binarize the grayscale image to obtain a binarized image, search the binarized image to obtain the upper edge point set and lower edge point set of the wafer, and take the point with the smallest vertical axis value in the lower edge point set as the position of the wafer notch center on the image; (4) Find straight lines in the grayscale image to obtain a straight line image; perform morphological closing operation on the straight lines in the straight line image, merge adjacent parallel lines to obtain horizontal fusion lines and vertical fusion lines. The width of the fusion lines is the grain gap width w1, and the intersection of the horizontal fusion lines and the vertical fusion lines is the grain corner point. Multi-target matching is performed on the grayscale image based on known grain corner points to obtain a series of grain corner points; (5) Based on the series of grain corner points obtained in step (4), calculate the average value of the horizontal adjacent corner points to obtain the grain width, calculate the average value of the vertical adjacent corner points to obtain the grain height, and calculate the grain deflection angle based on the position of the grain corner points. (6) Based on the data calculated in step (5), the column of grains closest to the position of the wafer notch center obtained in step (3) on the image is taken as the grain column on the vertical axis of the wafer; (7) The grain distribution of the entire wafer can be inferred from the position of the grains on the vertical axis of the wafer.

3. The image-based automatic grain positioning method according to claim 2, characterized in that: In step (3), a sliding window is set to move along the grayscale image, and the mean value within the sliding window is used as a threshold for binarization to obtain the binarized image.

4. The image-based automatic grain positioning method according to claim 3, characterized in that: Morphological operations are performed on the binarized image to denoise it.

5. The image-based automatic grain positioning method according to claim 2, characterized in that: In step (4), a grain corner is selected as the center on the grayscale image, and an image of size w2*w2 is used as the template image. Multi-target matching is performed in the grayscale image by the correlation coefficient method to obtain a series of grain corners, where w2>w1.

6. A computer device comprising a processor and a memory, the memory storing a computer program, the processor being configured to execute the computer program to implement the image-based automatic grain positioning method as described in any one of claims 1-5.

7. A computer storage medium storing a computer program, which, when executed on a processor, implements the image-based automatic die positioning method as described in any one of claims 1-5.

8. A computer program product comprising a computer program or instructions that, when executed by a processor, implement the image-based automatic die positioning method as described in any one of claims 1-5.

Images