Mould image processing method based on nanoimprint

By combining Z-shaped scanning and sliding window techniques with LBP encoding and Markov models to preprocess images of nanoimprint molds, the problem of noise interference in mold images is solved, and image quality and detection effect are improved.

CN118968090AActive Publication Date: 2024-11-15GERMANLITHO CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411458867.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-18
Publication Date
2024-11-15
Estimated Expiration
2044-10-18

AI Technical Summary

Technical Problem

During the nanoimprinting process, noise interference generated during mold image acquisition leads to a decrease in image quality, affecting the mold residue detection results.

Method used

The mold image is segmented using Z-shaped scanning and sliding window techniques, and preprocessed using LBP encoding and Markov model to generate texture transfer coefficients to smooth the image and improve image quality.

Benefits of technology

By extracting local texture features and edge information from mold images, the visualization effect of the images is improved, which facilitates subsequent mold residue detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118968090A_ABST
    Figure CN118968090A_ABST
Patent Text Reader

Abstract

The invention discloses a die image processing method based on nanoimprint, and belongs to the technical field of image processing, and the method comprises the following steps: S1, collecting a die processing image of nanoimprint, and traversing the die processing image to obtain a description visual operator; s2, preprocessing the mold processing image by using a description vision operator to obtain a to-be-monitored mold processing image; and S3, transmitting the to-be-monitored mold processing image to a user terminal. By analyzing the texture features, the visualization effect of the mold processing image can be improved, details in the image can be observed and recognized more easily, and subsequent residue detection and other processes can be conveniently carried out after the terminal receives the image.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of image processing, and in particular relates to a mold image processing method based on nanoimprinting. Background Art

[0002] Nanoimprint technology, as an advanced method for producing micro-nano patterns, has been widely used in the fields of semiconductor processing and optoelectronic device processing due to its high resolution, low cost and high throughput. However, in the nanoimprint process, due to the complexity of process parameters, the mold often has the problem of image residue after demolding, which seriously affects the quality of the mold in the subsequent processing process. The basis of the existing mold residue detection method is to collect images of the mold after the nanoimprint process is completed, and complete the mold residue detection through the image. However, a lot of interference will be generated in the existing image acquisition process, which makes the image noisy, resulting in a decrease in image quality and affecting the detection results. Summary of the invention

[0003] In order to solve the above problems, the present invention proposes a mold image processing method based on nanoimprinting.

[0004] The technical solution of the present invention is: a mold image processing method based on nanoimprinting comprises the following steps: S1, collecting the mold processing image of nanoimprinting, traversing the mold processing image, and obtaining a description visual operator; S2, preprocessing the mold processing image using a description visual operator to obtain the mold processing image to be monitored; S3. Transmit the mold processing image to be monitored to the user terminal.

[0005] Furthermore, S1 includes the following sub-steps: S11, collecting nanoimprint mold processing images; S12, using a sliding window to perform a Z-shaped scan on the mold processing image, dividing the mold processing image into a first visual image and a second visual image, and using pixel points obtained by the Z-shaped scan as a description pixel point set; S13, obtaining a corresponding first visual matrix and a second visual matrix according to the first visual image and the second visual image respectively; S14, obtaining a description solution based on the description pixel point set; S15. Using the first visual matrix, the second visual matrix and the description solution, a description visual operator is obtained.

[0006] The beneficial effect of the above further scheme is: in the present invention, the mold processing image is a rectangle, and there are two diagonals, which are referred to as the lower left diagonal (the upper right vertex and the lower left vertex are connected) and the lower right diagonal (the upper left vertex and the lower right vertex are connected) in the present invention. The present invention adopts a Z-shaped scan, so the lower left diagonal is involved. The mold processing image is scanned in a Z-shaped manner using a sliding window, that is, the sliding window traverses the mold processing image in the order of the first row from left to right, the lower left diagonal from top to bottom, and the last row from left to right, and the first row of pixels of the mold processing image, the pixels where the lower left diagonal is located, and the last row of pixels are used as a description pixel set. The lower left diagonal divides the mold processing image into two triangular areas, namely the first visual image and the second visual image. The square area is extracted from the triangular area to form a square matrix. The row and column values ​​of the square matrix can be used to calculate the description visual operator to reflect the color of the mold processing image.

[0007] Further, in S13, the largest square is extracted from the first visual image, and the pixel values ​​of each pixel in the largest square are arranged according to the pixel positions to obtain a first visual matrix; In S13, the largest square is extracted from the second visual image, and the pixel values ​​of each pixel in the largest square are arranged according to the pixel positions to obtain a second visual matrix.

[0008] The beneficial effect of the above further scheme is: in the present invention, since a square is extracted from a triangular-shaped visual image and the pixel values ​​of the square are arranged according to the pixel point positions, the visual matrix is ​​in the form of a square matrix, and the row and column values ​​of the visual matrix can be used for analysis and description.

[0009] Further, in S14, the calculation formula for solving M is described as: ; In the above, b represents the step size of the sliding window, and N represents the number of sliding times of the sliding window for Z-shaped scanning. It represents the mean pixel value of all pixels covered by the sliding window when it slides for the nth time. It represents the mean pixel value of all pixels covered by the sliding window when it slides for the n-1th time. represents the mean pixel value of all pixels covered by the sliding window when it slides for the n+1th time, sign(·) represents the sign function, and c represents a constant.

[0010] Furthermore, in S15, the calculation formula describing the visual operator a is: ; In the formula, ρ x represents the trace of the first visual matrix x, ρ y represents the trace of the second visual matrix y, M represents the description solution, Indicates rounding up.

[0011] Furthermore, S2 includes the following sub-steps: S21, performing LBP coding on each pixel point of the mold processing image to obtain a binary code of each pixel point, and converting the binary code into a decimal number to obtain an LBP value of each pixel point; S22, multiplying the description visual operator by the maximum grayscale value in the mold processing image to generate a texture description value; S23, calculating a texture migration coefficient for the mold processing image according to the texture description value and the LBP value of each pixel point; S24, using the texture migration coefficient to generate a mold processing image to be monitored.

[0012] The beneficial effect of the above further scheme is: in the present invention, the basic idea of ​​LBP coding is to describe the local texture information of the image by comparing the grayscale value of each pixel with its surrounding pixels and converting the comparison result into a binary number, which can flexibly describe the local texture characteristics of the image.

[0013] Furthermore, in S3, the calculation formula of the texture migration coefficient W is: ; Where, LBP k Indicates the LBP value of the k-th pixel, LBP k+1 represents the LBP value of the k+1th pixel, P represents the texture description value, and K represents the total number of pixels in the mold processing image.

[0014] Furthermore, in S24, the texture migration coefficient is used as a smoothing coefficient of the Markov model, and the mold processing image is smoothed using the Markov model to obtain the mold processing image to be monitored.

[0015] The beneficial effect of the above further solution is that in the present invention, the smoothing coefficient of the Markov model can help the model better adapt to the changing trend of the image data, thereby improving the stability and prediction accuracy of the model. The Markov model can smooth out the noise and outliers in the image data, making the image smoother and more stable.

[0016] The beneficial effects of the present invention are as follows: the present invention can traverse and preprocess the mold processing image, and use the pixel point features and texture features of the mold processing image to complete image analysis, so that the final mold processing image to be monitored can more prominently highlight the texture features in the image, and make the information and edges in the image clearer; by analyzing the texture features, the visualization effect of the mold processing image can be improved, making the details in the image easier to observe and identify, and facilitating subsequent residue detection and other processes after the terminal receives the image. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Flow chart of the mold image processing method based on nanoimprinting. DETAILED DESCRIPTION

[0018] The embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

[0019] like Figure 1 As shown, the present invention provides a mold image processing method based on nanoimprinting, comprising the following steps: S1, collecting the mold processing image of nanoimprinting, traversing the mold processing image, and obtaining a description visual operator; S2, preprocessing the mold processing image using a description visual operator to obtain the mold processing image to be monitored; S3. Transmit the mold processing image to be monitored to the user terminal.

[0020] In this embodiment of the present invention, S1 includes the following sub-steps: S11, collecting nanoimprint mold processing images; S12, using a sliding window to perform a Z-shaped scan on the mold processing image, dividing the mold processing image into a first visual image and a second visual image, and using pixel points obtained by the Z-shaped scan as a description pixel point set; S13, obtaining a corresponding first visual matrix and a second visual matrix according to the first visual image and the second visual image respectively; S14, obtaining a description solution based on the description pixel point set; S15. Using the first visual matrix, the second visual matrix and the description solution, a description visual operator is obtained.

[0021] In the present invention, the mold processing image is a rectangle with two diagonals, which are referred to as the lower left diagonal (connecting the upper right vertex and the lower left vertex) and the lower right diagonal (connecting the upper left vertex and the lower right vertex) in the present invention. The present invention adopts a Z-shaped scan, so the lower left diagonal is involved. The mold processing image is scanned in a Z-shaped manner using a sliding window, that is, the sliding window traverses the mold processing image in the order of the first row from left to right, the lower left diagonal from top to bottom, and the last row from left to right, and the first row of pixels of the mold processing image, the pixels where the lower left diagonal is located, and the last row of pixels are used as a description pixel set. The lower left diagonal divides the mold processing image into two triangular areas, namely the first visual image and the second visual image. A square area is extracted from the triangular area to form a square matrix. The row and column values ​​of the square matrix can be used to calculate the description visual operator to reflect the color of the mold processing image.

[0022] In the embodiment of the present invention, in S13, the largest square is extracted from the first visual image, and the pixel values ​​of each pixel in the largest square are arranged according to the pixel positions to obtain a first visual matrix; In S13, the largest square is extracted from the second visual image, and the pixel values ​​of each pixel in the largest square are arranged according to the pixel positions to obtain a second visual matrix.

[0023] In the present invention, since a square is extracted from a triangular-shaped visual image and the pixel values ​​of the square are arranged according to the pixel point positions, the visual matrix is ​​in the form of a square matrix, and the row and column values ​​of the visual matrix can be used for analysis and description.

[0024] In the embodiment of the present invention, in S14, the calculation formula for calculating M is described as follows: ; In the above, b represents the step size of the sliding window, and N represents the number of sliding times of the sliding window for Z-shaped scanning. It represents the mean pixel value of all pixels covered by the sliding window when it slides for the nth time. It represents the mean pixel value of all pixels covered by the sliding window when it slides for the n-1th time. represents the mean pixel value of all pixels covered by the sliding window when it slides for the n+1th time, sign(·) represents the sign function, and c represents a constant.

[0025] In the embodiment of the present invention, in S15, the calculation formula describing the visual operator a is: ; In the formula, ρ x represents the trace of the first visual matrix x, ρ y represents the trace of the second visual matrix y, M represents the description solution, Indicates rounding up.

[0026] In this embodiment of the present invention, S2 includes the following sub-steps: S21, performing LBP coding on each pixel point of the mold processing image to obtain a binary code of each pixel point, and converting the binary code into a decimal number to obtain an LBP value of each pixel point; S22, multiplying the description visual operator by the maximum grayscale value in the mold processing image to generate a texture description value; S23, calculating a texture migration coefficient for the mold processing image according to the texture description value and the LBP value of each pixel point; S24, using the texture migration coefficient to generate a mold processing image to be monitored.

[0027] In the present invention, the basic idea of ​​LBP coding is to describe the local texture information of the image by comparing the grayscale values ​​of each pixel with its surrounding pixels and converting the comparison result into a binary number, which can flexibly describe the local texture features of the image.

[0028] In the embodiment of the present invention, in S3, the calculation formula of the texture migration coefficient W is: ; Where, LBP k Indicates the LBP value of the k-th pixel, LBP k+1 represents the LBP value of the k+1th pixel, P represents the texture description value, and K represents the total number of pixels in the mold processing image.

[0029] In the embodiment of the present invention, in S24, the texture migration coefficient is used as the smoothing coefficient of the Markov model, and the mold processing image is smoothed using the Markov model to obtain the mold processing image to be monitored.

[0030] In the present invention, the smoothing coefficient of the Markov model can help the model better adapt to the changing trend of the image data, thereby improving the stability and prediction accuracy of the model. The Markov model can smooth out the noise and outliers in the image data, making the image smoother and more stable.

[0031] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A mold image processing method based on nanoimprinting, characterized in that: The following steps are involved: S1, collecting the mold processing image of nanoimprinting, traversing the mold processing image, and obtaining a description visual operator; S2, preprocessing the mold processing image using a description visual operator to obtain the mold processing image to be monitored; S3. Transmit the mold processing image to be monitored to the user terminal.

2. The mold image processing method based on nanoimprinting according to claim 1, characterized in that: The S1 comprises the following sub-steps: S11, collecting nanoimprint mold processing images; S12, using a sliding window to perform a Z-shaped scan on the mold processing image, dividing the mold processing image into a first visual image and a second visual image, and using pixel points obtained by the Z-shaped scan as a description pixel point set; S13, obtaining a corresponding first visual matrix and a second visual matrix according to the first visual image and the second visual image respectively; S14, obtaining a description solution based on the description pixel point set; S15. Using the first visual matrix, the second visual matrix and the description solution, a description visual operator is obtained.

3. The mold image processing method based on nanoimprinting according to claim 2, characterized in that: In S13, the largest square is extracted from the first visual image, and the pixel values ​​of each pixel in the largest square are arranged according to the pixel positions to obtain a first visual matrix; In S13, the largest square is extracted from the second visual image, and the pixel values ​​of each pixel in the largest square are arranged according to the pixel positions to obtain a second visual matrix.

4. The mold image processing method based on nanoimprinting according to claim 2, characterized in that: In S14, the calculation formula for solving M is described as follows: ; In the equation, b represents the step size of the sliding window, and N represents the number of times the sliding window performs a Z-shaped scan. It represents the mean pixel value of all pixels covered by the sliding window when it slides for the nth time. It represents the mean pixel value of all pixels covered by the sliding window when it slides for the n-1th time. represents the mean pixel value of all pixels covered by the sliding window when it slides for the n+1th time, sign(·) represents the sign function, and c represents a constant.

5. The mold image processing method based on nanoimprinting according to claim 2, characterized in that: In S15, the calculation formula describing the visual operator a is: ; In the formula, ρ x represents the trace of the first visual matrix x, ρ y represents the trace of the second visual matrix y, M represents the description solution, Indicates rounding up.

6. The mold image processing method based on nanoimprinting according to claim 1, characterized in that: The S2 comprises the following sub-steps: S21, performing LBP coding on each pixel point of the mold processing image to obtain a binary code of each pixel point, and converting the binary code into a decimal number to obtain an LBP value of each pixel point; S22, multiplying the description visual operator by the maximum grayscale value in the mold processing image to generate a texture description value; S23, calculating a texture migration coefficient for the mold processing image according to the texture description value and the LBP value of each pixel point; S24, using the texture migration coefficient to generate a mold processing image to be monitored.

7. The mold image processing method based on nanoimprinting according to claim 6, characterized in that: In S3, the calculation formula of the texture migration coefficient W is: ; Where, LBP k Indicates the LBP value of the k-th pixel, LBP k+1 represents the LBP value of the k+1th pixel, P represents the texture description value, and K represents the total number of pixels in the mold processing image.

8. The mold image processing method based on nanoimprinting according to claim 6, characterized in that: In S24, the texture migration coefficient is used as a smoothing coefficient of the Markov model, and the mold processing image is smoothed using the Markov model to obtain the mold processing image to be monitored.

Citation Information

Patent Citations

  • Method for removing speckles from SAR (Synthetic Aperture Radar) image based on MRF (Markov Random Field) priority

    CN102136134A

  • Video abstraction method and device based on dual-mode characteristics and attention mechanism

    CN113158720A

  • Improved texture mapping method and device based on Markov random field

    CN115205435A

  • Method for describing uniform texture and its descriptor

    JP2001266161A