A parametric curve mask OPC method, device, medium, and product based on piecewise Bezier.

Abstract

This application discloses a parametric curve mask OPC method, apparatus, medium, and product based on piecewise Bézier curves, relating to the field of semiconductor manufacturing technology. The method includes: performing continuous transfer mask optimization on a target pattern to obtain a CTM mask with auxiliary patterns; then performing level set optimization to obtain a mask layout; after thinning the contour of the mask layout, fitting it with a Bézier curve to obtain a pixelated mask; then constructing a mask error factor matrix with edge placement error as the target, and solving for the movement of control points; adjusting the positions of the control points according to the angle bisector direction based on the movement of the control points to obtain new control points; fitting the new control points with a Bézier curve, and calculating the weighted edge placement error of the evaluation points in the fitting result. This application can reduce data volume while ensuring imaging quality.

Description

Technical Field

[0001] This application relates to the field of semiconductor manufacturing technology, and in particular to a parametric curve mask OPC method, apparatus, medium, and product based on segmented Bezier. Background Technology

[0002] With the rapid development of semiconductor manufacturing technology, integrated circuit design has become increasingly complex, leading to a surge in data volume for Graphic Data System II (GDSII) or Open Artwork System Interchange Standard (OASIS). Furthermore, curve masks based on Inverse Lithography Technology (ILT) introduce more complex free-form SRAFs, whose data volume becomes more difficult to control compared to traditional Manhattan layouts. This data consumes significant storage resources and bandwidth, resulting in substantial resource and time costs.

[0003] The advent of multi-beam mask writers has made the large-scale application of curve masks possible, but the whole-chip curve mask ecosystem still faces many challenges. In curve mask technology, ILT (In-line Transformer) is pixel-based. Compared to traditional edge-based optical proximity correction (OPC), most existing algorithms result in a greater number of optimization variables during layout optimization. Furthermore, pixel-based or curve contour-based layout data is either difficult to format using existing data formats or its file size increases dramatically, making layout storage format increasingly important in curve mask data storage. Therefore, effectively reducing mask data storage while maintaining image quality is crucial for improving production efficiency, ensuring product quality, and guaranteeing manufacturability. It is also a significant driver of technological innovation in the semiconductor industry. Summary of the Invention

[0004] The purpose of this application is to provide a parametric curve mask OPC method, device, medium, and product based on piecewise Bezier, which can reduce the amount of data while ensuring imaging quality.

[0005] To achieve the above objectives, this application provides the following solution:

[0006] Firstly, this application provides a parametric curve mask OPC method based on piecewise Bézier curves, including:

[0007] Obtain the target image;

[0008] The target graphic is subjected to continuous transmission mask optimization to obtain a CTM mask with auxiliary graphics;

[0009] Horizontal set optimization is performed on the CTM mask with auxiliary graphics to obtain the mask layout;

[0010] The outline of the mask pattern is thinned, and the thinning points are fitted with a Bézier curve to obtain a pixelated mask; the pixelated mask includes auxiliary graphics and main graphics;

[0011] Evaluation points are selected on the outline of the target graphic;

[0012] Based on the evaluation points and control points, a mask error factor matrix is ​​constructed with edge placement error as the target; the control points are either thinned points on the main graphic or points obtained by selecting every k points on the outline of the main graphic.

[0013] The movement of the control points is obtained by multiplying the mask error factor matrix by the diagonal weighting matrix.

[0014] Based on the amount of movement of the control point, adjust the position of the control point according to the angle bisector direction to obtain a new control point; the angle bisector direction is the direction of the angle bisector of the angle formed by the current control point and the two adjacent control points.

[0015] Perform Bézier curve fitting on the new control points and calculate the weighted edge placement error of the evaluation points in the fitting results.

[0016] Optionally, the target pattern is subjected to continuous transmission mask optimization to obtain a CTM mask with auxiliary patterns, specifically including:

[0017] The mode error is used as the loss function for continuous transmission mask optimization, and sub-resolution auxiliary patterns are extracted to obtain a CTM mask with auxiliary patterns.

[0018] The loss function formula is:

[0019] Among them, F peloss Let m and n be the number of rows and columns of the target image, respectively, and i and j be the row index and column index, respectively. z is the pixel value of the target graphic. i,j This represents the pixel value of the wafer image.

[0020] Optionally, the step of performing level set optimization on the CTM mask with auxiliary patterns to obtain the mask layout specifically includes:

[0021] The CTM mask with auxiliary graphics is adaptively binarized to obtain the initial value mask;

[0022] The initial mask is optimized by level set optimization to obtain the mask layout.

[0023] Optionally, the contour of the mask layout is thinned, and the thinned points are fitted with a Bézier curve to obtain a pixelated mask, specifically including:

[0024] The outline of the mask pattern is thinned, and the thinned points are fitted with a Bézier curve to obtain the fitted curve.

[0025] The fitted curve is rendered as a pixelated mask using a multisampling method.

[0026] Optionally, the movement of the control points can be obtained by multiplying the mask error factor matrix by a diagonal weighting matrix, specifically including:

[0027] The weighted mask error factor matrix is ​​obtained by multiplying the mask error factor matrix by the diagonal weighting matrix.

[0028] The control point movement is obtained by solving the weighted mask error factor matrix using the SVD truncation method. The calculation formula is as follows:

[0029] e≈e0+M·d

[0030]

[0031] Where e0 is the initial edge placement error, represented as an n×1 column vector; n is the number of selected evaluation points; M is the mask error factor matrix; d is the amount of control point movement; and e is the edge placement error corresponding to the selected evaluation point on the target graphic after adjusting the control point movement.

[0032] The MEEF matrix structure is as follows:

[0033]

[0034] Where e1, e2, ..., e r The edge placement error for each evaluation point is d1, d2, ..., d. n Let r be the movement amount of each control point, r be the total number of evaluation points, and n be the total number of control points.

[0035] Optionally, Bézier curves are fitted to the new control points, and the weighted edge placement error of the evaluation points in the fitting result is calculated, specifically including:

[0036] The new control points are fitted with Bézier curves to obtain the fitting results.

[0037] The edge placement error of the midpoint evaluation point on each side is calculated based on the fitting results to obtain the weighted edge placement error.

[0038] Optionally, after performing Bézier curve fitting on the new control points and calculating the weighted edge placement error of the evaluation points in the fitting result, the method further includes:

[0039] Using the new control point as the control point and the weighted edge placement error as the edge placement error, return to the step "Construct a mask error factor matrix based on the evaluation point and control point, with the edge placement error as the target" until the weighted edge placement error of all evaluation points is less than the preset threshold or the number of iterations is reached.

[0040] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the parametric curve mask OPC method based on piecewise Bézier as described above.

[0041] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the parametric curve mask OPC method based on piecewise Bézier described above.

[0042] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the piecewise Bézier-based parametric curve mask OPC method described above.

[0043] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0044] This application provides a parametric curve mask OPC method, device, medium, and product based on piecewise Bézier curves. The method includes: acquiring a target image; performing continuous transmission mask optimization on the target image to obtain a CTM mask with auxiliary images; performing level set optimization on the CTM mask with auxiliary images to obtain a mask layout; thinning the contour of the mask layout and fitting the thinned points with a Bézier curve to obtain a pixelated mask; the pixelated mask includes auxiliary images and a main image; selecting evaluation points on the contour of the target image; and, based on the evaluation points and control points, using edge... Using placement error as the target, a mask error factor matrix is ​​constructed. The control points are either thinned points on the main image or points obtained by selecting every k points on the main image contour. The movement of the control points is obtained by multiplying the mask error factor matrix by a diagonal weighted matrix. Based on the movement of the control points, the positions of the control points are adjusted according to the angle bisector direction to obtain new control points. The angle bisector direction is the direction of the angle bisector of the angle formed by the current control point and its two adjacent control points. Bézier curves are fitted to the new control points, and the weighted edge placement error of the evaluation points in the fitting result is calculated. This application reduces the data of the curve mask by thinning the mask contour and fitting it with Bézier curves, and optimizes the control points on the main image using an optimization algorithm based on the mask error factor matrix. This reduces the amount of data without degrading the image quality. Therefore, this application can ensure image quality while reducing the amount of data. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is an application environment diagram of a parametric curve mask OPC method based on piecewise Bezier curves in one embodiment of this application.

[0047] Figure 2 This is a flowchart illustrating a parametric curve mask OPC method based on piecewise Bezier curves, provided as an embodiment of this application.

[0048] Figure 3 This is a schematic diagram of the target pattern, the mask layout after level set optimization, and the wafer image result of the mask provided in an embodiment of this application.

[0049] Figure 4 This is a schematic diagram showing the selection of initial master graphic control points and Ep points for different thinning thresholds provided in an embodiment of this application.

[0050] Figure 5 This is a schematic diagram of a mask layout optimized with different thresholds and the resulting wafer image of the mask, provided as an embodiment of this application.

[0051] Figure 6 This is a schematic diagram of the changes in EPE and wEPE during the optimization process provided in an embodiment of this application.

[0052] Figure 7 This is a schematic diagram showing the number of contour points under different threshold optimizations provided in an embodiment of this application.

[0053] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0054] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0055] This application provides a parametric curve mask optimization method based on piecewise Bézier, belonging to the OPC method and mask data compression in computational lithography. Compared with the traditional optical proximity correction (OPC) method for linear masks, this application is an OPC method for parametric curve masks. This application first performs continuous transmission mask (CTM) optimization on the target graphic to obtain a mask with sub-resolution assist features (SRAFs). Second, the CTM mask is adaptively binarized and used as the initial value for subsequent level set optimization. Then, level set optimization is performed on the initial mask to obtain the optimized mask layout. Next, the layout outline is thinned. While maintaining the overall shape characteristics of the curve, key points are selected and redundant points are removed using an algorithm, thereby reducing the number of points on the outline and simplifying the data. The thinned points are used as control points, fitted with piecewise Bézier curves, and the fitted outline is re-rendered as a pixelated layout, which is then used as the initial layout for optimization. Then, the fitted SRAFs are fixed, and a Mask Error Enhancement Factor (MEEF) matrix is ​​constructed with edge placement error (EPE) as the target. The movement of the main graphic control points is obtained by solving the matrix equation. The fitting, rendering, and MEEF matrix construction operations are performed again for iterative optimization until the number of iterations or the error is less than the tolerance. This application can effectively reduce the amount of OPC layout data while ensuring imaging quality and process requirements.

[0056] Storing pixel-based layout data in GDSII format leads to a significant increase in data file size, and GDSII format data for curve masks is even larger. Therefore, a new layout storage format is needed to reduce data volume while maintaining image quality. Compared to existing methods, this method selects mask control points and uses Bézier curve fitting. By optimizing the position of the control points, it can effectively reduce the amount of mask data stored while maintaining image quality. This method represents the mask contour using piecewise Bézier curves, requiring only the parameters of the corresponding curves to be saved, which can significantly reduce the amount of mask data stored after level set optimization.

[0057] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0058] The parametric curve mask OPC method based on piecewise Bézier provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on other servers. Terminal 102 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Server 104 can be implemented using a standalone server, a server cluster consisting of multiple servers, or a cloud server.

[0059] In one exemplary embodiment, such as Figure 2 As shown, a parametric curve mask OPC method based on piecewise Bézier is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps S1 to S9.

[0060] in:

[0061] S1. Obtain the target image. Read the target image and use it as the initial image for CTM optimization. The target image is a pixel image of size x*x (x is an integer) with values ​​of 0 or 1. For example, a target image of size 257*257 pixels where the transparent part is 1 and the opaque part is 0. Please refer to [link / reference]. Figure 3 .

[0062] S2. Perform continuous transmission mask optimization on the target graphic to obtain a CTM mask with auxiliary graphics.

[0063] In this embodiment, pattern error (PE) is selected as the loss function for both CTM and level set methods.

[0064] The mode error is used as the loss function for continuous transmission mask optimization, and sub-resolution auxiliary patterns are extracted to obtain a CTM mask with auxiliary patterns.

[0065] The loss function formula is:

[0066] Among them, F peloss Let m and n be the number of rows and columns of the target image, respectively, and i and j be the row index and column index, respectively. z is the pixel value of the target graphic. i,jThis represents the pixel value of the wafer image.

[0067] CTM optimization relaxes the mask pixels from only 0 and 1 to a range of 0 to 1, reducing the optimization difficulty and enabling the generation of SRAFs (Supporting Graphs). However, CTM masks are not fabricable and require intensity quantization.

[0068] S3. Perform level set optimization on the CTM mask with auxiliary graphics to obtain the mask layout.

[0069] In this embodiment, the CTM mask with auxiliary graphics (SRAFs) is adaptively binarized and used as the initial value for subsequent level set optimization. Other methods for obtaining initial values ​​include rule-based, model-based, and fixed-width curve SARF placement. In this embodiment, the initial value for level set optimization is obtained through CTM, and level set optimization is then performed. The level set optimization method represents the two-dimensional mask as the zero level set of the three-dimensional level set function, thereby optimizing the objective function by moving the mask contour (by moving it in the gradient direction of the level set function), which is then transformed into optimizing the mask contour through the evolution of the level set function. In this embodiment, the gradient of optimization is calculated through variational derivation, and when calculating the partial differential equation of the level set function, Weighted Essentially Non-Oscillatory (WENO) spatial discretization and Total Variation Diminishing-Runge-Kutta (TVD-RK) time discretization are applied.

[0070] The level set method represents the mask profile as the zero level set of a level set function that is one dimension higher than the initial mask (obtained by adaptive binarization of a CTM mask with auxiliary graphics). First, the mask profile is implicitly represented as the zero level set of the level set function Φ(r,t), the evolution of which is governed by the Hamilton-Jacobi partial differential equation: Secondly, when solving partial differential equations, the following methods are used:

[0071] The solution is obtained using WENO spatial discretization and TVD-RK time discretization. Furthermore, it is achieved through... The discretization rule updates the level set function, where... The level set function represents the level set function at the (n+1)th iteration. V represents the level set function at the nth iteration, where Δt is the iteration time determined by the CFL condition. n Indicates the contour movement speed. Let represent the gradient of the level set function at the nth iteration. The zero level set of the updated level set function is the optimized mask. The initialization of the level set function is determined by... Preprocessing is performed, in which It is the mask after CTM binarization, where δ is roughly equivalent to the spatial step size. Through Initialization is performed through iterative updates. After initialization, level set optimization is performed using mask gradient information. The iterative formula is: Where V m Indicates the speed at which the mask contour moves. This is the curvature regularization term.

[0072] S4. Thin out the outline of the mask pattern and fit the thinned points with a Bézier curve to obtain a pixelated mask; the pixelated mask includes auxiliary graphics and main graphics.

[0073] The contour of the mask layout after level set optimization is thinned. While maintaining the overall shape characteristics of the curve, key points are selected and redundant points are removed through algorithms, thereby reducing the number of data points on the contour and simplifying the data. The thinned points are then fitted with a Bézier curve, and the fitted curve is rendered into a pixelated mask using Multi-Sample Anti-Aliasing (MSAA). Figure 4 As shown, in this embodiment, four different thresholds were selected, namely 0.7 ( Figure 4 (a)), 0.8 ( Figure 4 (b) ), 0.9 ( Figure 4 (c)), 1.0 ( Figure 4 (d) is the initial main graphic control point. Figure 4 (e) is the Ep point selection diagram.

[0074] The specific screening process includes: starting from the endpoints of the contour and connecting them, continuously calculating the vertical distance from all points to the line segment and finding the farthest point. If the distance is greater than a threshold, the point is retained and the segmented contour is processed recursively; otherwise, the intermediate points are discarded. This process is used to screen key points and eliminate redundant points, thereby reducing the number of points on the contour.

[0075] S5. Select evaluation points on the contour of the target graphic. In this embodiment, evaluation points (Ep) are selected on the target graphic for subsequent mask optimization. The optimization results are as follows: Figure 5 As shown, Figure 5 In the diagram, (a1) and (a2) represent the optimized mask layout and the wafer image of the mask when the thinning threshold is 0.7, respectively; (b1) and (b2) represent the optimized mask layout and the wafer image of the mask when the thinning threshold is 0.8, respectively; (c1) and (c2) represent the optimized mask layout and the wafer image of the mask when the thinning threshold is 0.9, respectively; and (d1) and (d2) represent the optimized mask layout and the wafer image of the mask when the thinning threshold is 1.0, respectively.

[0076] S6. Based on the evaluation points and control points, construct a mask error factor matrix with edge placement error as the target; the control points are either thinning points on the main graphic or points obtained by selecting every k points on the outline of the main graphic.

[0077] The fitted SRAFs (auxiliary graphics in the pixelated mask) are fixed and remain unchanged during the optimization process. The sparse points on the main graphic are used as control points, or points are taken every k points on the contour as control points. The MEEF matrix is ​​constructed through edge placement error (EPE), and the movement of the control points is obtained by solving the matrix equation.

[0078] S7. Solve by multiplying the mask error factor matrix by the diagonal weighted matrix to obtain the movement of the control points.

[0079] First, the mask error factor matrix is ​​multiplied by the diagonal weighting matrix to obtain the weighted mask error factor matrix; then, the SVD truncation method is used to solve the weighted mask error factor matrix to obtain the movement of the control points.

[0080] Because the sparsity of the MEEF matrix results in a large condition number, it is very sensitive to changes in the input and cannot be solved by conventional methods. In this method, we use the Singular Value Decomposition (SVD) truncation method to remove smaller singular values, effectively filtering out the noise amplification effect caused by small singular values ​​and improving the stability of the system.

[0081] The EPE calculation formula is as follows:

[0082]

[0083] Among them, F epeloss Let m and n be the objective function, representing the number of rows and columns of the target graph, respectively, and i and j be the row and column indices, respectively. i,j tr represents the spatial image pixel value, and tr represents the intensity threshold. This represents the spatial image gradient.

[0084] The formula for solving the MEEF matrix is ​​as follows:

[0085] e≈e0+M·d

[0086]

[0087] The first formula here approximates the relationship between the initial error and the error after the movement amount d. Therefore, in this embodiment, the error after movement is set to 0. Solving the matrix equation yields the movement amount d that makes the error zero, thus reducing the error by moving the control points. The solution process is as follows: 1: Calculate the initial error and construct the MEEF matrix. 2: Solve the matrix equation to obtain the movement amount d. Where e0 is the initial edge placement error, represented as an n×1 column vector; n is the number of selected evaluation points; M is the mask error factor matrix; d is the movement amount of the control points; and e is the edge placement error corresponding to the selected evaluation points on the target graphic after adjusting the control point movement amount.

[0088] The MEEF matrix structure is as follows:

[0089]

[0090] Where e1, e2, ..., e r The edge placement error for each evaluation point is d1, d2, ..., d. n Let r be the movement amount of each control point, r be the total number of evaluation points, and n be the total number of control points.

[0091] S8. Based on the movement of the control point, adjust the position of the control point according to the angle bisector direction to obtain a new control point; the angle bisector direction is the direction of the angle bisector of the angle formed by the current control point and its two adjacent control points.

[0092] The movement amount of the control point is obtained through step S7. It can be moved in any direction to obtain a new control point. In this embodiment, the movement direction is the direction of the angle bisector of the angle formed by the current control point and its two adjacent control points.

[0093] S9. Fit the new control points to a Bézier curve and calculate the weighted edge placement error of the evaluation points in the fitting results.

[0094] In this embodiment, the new control points are first fitted with Bézier curves to obtain the fitting results; then, the edge placement error of the midpoint evaluation point on each side is calculated based on the fitting results to obtain the weighted edge placement error.

[0095] The new control points are fitted with Bézier curves, rendered, and their EPE (Evaluation Point Error) is calculated for all evaluation points. A diagonal weighted matrix W is introduced in different regions to assign differentiated weights to the variables. Specifically, to enhance the optimization focus effect, the weight values ​​of the midpoints of each edge of the target graphic are specifically increased. The improved EPE of each edge midpoint is called wEPE (weighted edgeplace error). The weighted matrix allows different weights to be assigned to different ep points during the optimization process, causing optimization resources to be tilted towards the ep points of interest. For example, in this embodiment, the ep points of the four edge midpoints are optimized.

[0096] Finally, using the new control point as the control point and the weighted edge placement error as the edge placement error, return to the step "Construct a mask error factor matrix based on the evaluation points and control points, with the edge placement error as the target" until the weighted edge placement error of all evaluation points is less than the preset threshold or the number of iterations is reached. That is, repeat steps S6-S9 until the number of optimizations is reached or the EPE and wEPE errors meet the requirements. After optimization, the grayscale mask is obtained, and the result is as follows. Figure 3 As shown ( Figure 3 (a) is the target graphic. Figure 3 (b) Mask layout after level set optimization Figure 3 (c) is the wafer image result of this mask.

[0097] In addition, please see Figure 6 and Figure 7 , Figure 6 The curves showing the changes in EPE and wEPE during the optimization process are illustrated ((a) shows the change in EPE, and (b) shows the change in wEPE). After optimization of the initial level set, the EPE before thinning and fitting was 7.7184, and the wEPE was 1.5376. After 50 iterations of optimization, the error results under different thinning thresholds are as follows: when the threshold is 0.7, the EPE decreases to 2.1292, and the wEPE decreases to 0.000113; when the threshold is 0.8, the EPE is 2.1508, and the wEPE is 0.000099; when the threshold is 0.9, the EPE is 1.6505, and the wEPE is 0.000128; when the threshold is 1.0, the EPE further decreases to 1.5440, and the wEPE is 0.000107. The results show that the EPE at all thresholds is significantly lower than the initial value, and the wEPE is almost zero, indicating that the midpoints of the key edges in the wafer image accurately fall on the target contour, fully meeting the requirements of industrial-grade image accuracy, and verifying the effectiveness of this method in maintaining imaging accuracy.

[0098] Figure 7The number of contour points was optimized for different thresholds. After level set optimization, the original number of contour points was 4944. When the thinning thresholds were set to 0.7, 0.8, 0.9, and 1.0, the number of contour points decreased to 881, 795, 708, and 649, respectively. This demonstrates that the proposed method can significantly compress the mask contour data while maintaining wafer image quality, achieving nearly 80% contour point compression and effectively reducing data processing overhead in the OPC and manufacturing stages.

[0099] The principle of this method is as follows: Based on the target image, firstly, the initial mask image is optimized using the CTM algorithm to obtain a grayscale mask containing SRAFs. This optimization process based on the CTM algorithm can improve image quality, but CTMs are not manufacturable. The grayscale mask is binarized and used as the initial value for the level set method optimization. Then, the mask is represented by the level set algorithm as the zero level set of a level set function that is one dimension higher than the original mask. The optimization of the objective function by moving the mask contour is transformed into optimization through the evolution of the level set function, resulting in a mask image with high imaging accuracy. Using traditional data formats to store layout contour data points leads to excessively large layout files, which are difficult for current industries to process. Next, the mask contour is thinned and each segment is fitted with a Bézier curve to reduce the data of the curve mask. A multisampling method is used to render the fitted contour into a pixelated mask. The SRAFs are fixed, and the thinned points (control points) on the main image are optimized using an optimization algorithm based on the MEEF matrix. This algorithm can reduce the amount of data without degrading the image quality.

[0100] This embodiment provides a parametric curve mask optimization method based on piecewise Bézier curves. First, the target mask image is optimized using the CTM algorithm to obtain a grayscale mask containing SRAFs. This CTM-based optimization process ensures image quality but is not manufacturable and requires intensity quantization optimization. The grayscale mask is binarized and used as the initial value for level set optimization. Then, the mask is represented as the zero level set of a level set function one dimension higher than the original mask using a level set algorithm. The optimization of the objective function is transformed from moving the mask contour to achieving optimization through the evolution of the level set function, resulting in an optimized mask layout contour. At this point, storing the layout contour data points using traditional data formats leads to excessively large layout files, which are difficult for current industries to process. Finally, the mask contour is thinned and fitted with a Bézier curve to reduce the data of the curve mask contour, and a multisampling method is used to render the fitted contour as a pixelated mask. With SRAFs fixed, the thinned points (control points) on the main image are optimized using a MEEF matrix-based optimization algorithm. This algorithm can preserve SRAF details without degrading image quality while reducing the amount of data.

[0101] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 8 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a piecewise Bézier-based parametric curve mask OPC method.

[0102] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0103] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0104] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0105] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0106] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0107] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this 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), magnetic 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. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0108] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0109] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above 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.

[0110] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A parametric curve mask OPC method based on piecewise Bézier curves, characterized in that, include: Obtain the target image; The target graphic is subjected to continuous transmission mask optimization to obtain a CTM mask with auxiliary graphics; The CTM mask is a grayscale mask obtained by optimizing a continuous transmission mask. Horizontal set optimization is performed on the CTM mask with auxiliary graphics to obtain the mask layout; The outline of the mask pattern is thinned, and the thinned points are fitted with a Bézier curve to obtain a pixelated mask. The pixelated mask includes auxiliary graphics and a main graphic; The auxiliary graphics in the pixelated mask are fixed and remain unchanged in subsequent optimization processes; Evaluation points are selected on the outline of the target graphic; Based on the evaluation points and control points, a mask error factor matrix is ​​constructed with edge placement error as the target; the control points are either thinned points on the main graphic or points obtained by selecting every k points on the outline of the main graphic. The movement of the control points is obtained by multiplying the mask error factor matrix by the diagonal weighting matrix. Based on the amount of movement of the control point, adjust the position of the control point according to the angle bisector direction to obtain a new control point; the angle bisector direction is the direction of the angle bisector of the angle formed by the current control point and the two adjacent control points. Fit the new control points to a Bézier curve and calculate the weighted edge placement error of the evaluation points in the fitting results; The target pattern is subjected to continuous transmission mask optimization to obtain a CTM mask with auxiliary patterns, specifically including: The mode error is used as the loss function for continuous transmission mask optimization, and sub-resolution auxiliary patterns are extracted to obtain a CTM mask with auxiliary patterns. The loss function formula is: , in, For loss function, and These represent the number of rows and columns of the target graphic, respectively. and These are row indexes and column indexes, respectively. For the target image pixel value, This represents the pixel value of the wafer image.

2. The parametric curve mask OPC method based on piecewise Bézier as described in claim 1, characterized in that, The step of performing level set optimization on the CTM mask with auxiliary graphics to obtain the mask layout specifically includes: The CTM mask with auxiliary graphics is adaptively binarized to obtain the initial value mask; The initial mask is optimized by level set optimization to obtain the mask layout.

3. The parametric curve mask OPC method based on piecewise Bézier as described in claim 1, characterized in that, The contour of the mask layout is thinned, and the thinned points are fitted with a Bézier curve to obtain a pixelated mask, specifically including: The outline of the mask pattern is thinned, and the thinned points are fitted with a Bézier curve to obtain the fitted curve. The fitted curve is rendered as a pixelated mask using a multisampling method.

4. The OPC method for parametric curve masks based on piecewise Bézier as described in claim 1, characterized in that, The movement of the control points is obtained by multiplying the mask error factor matrix by a diagonal weighting matrix and then solving the solution. Specifically, this includes: The weighted mask error factor matrix is ​​obtained by multiplying the mask error factor matrix by the diagonal weighting matrix. The movement of the control points is obtained by solving the weighted mask error factor matrix using the singular value decomposition truncation method. The calculation formula is as follows: in, The initial edge placement error is expressed as Column vectors; The number of selected evaluation points; This is the mask error factor matrix; denoted as the amount of movement of the control point; e represents the edge placement error corresponding to the selected evaluation point on the target graphic after adjusting the amount of movement of the control point. The structure of the mask error factor matrix is ​​as follows: ； Among them, e1, e2, ..., e r The edge placement error for each evaluation point is d1, d2, ..., d. n Let r be the movement amount of each control point, r be the total number of evaluation points, and n be the total number of control points.

5. The OPC method for parametric curve masks based on piecewise Bézier as described in claim 1, characterized in that, Perform Bézier curve fitting on the new control points and calculate the weighted edge placement error of the evaluation points in the fitting results, specifically including: The new control points are fitted with Bézier curves to obtain the fitting results. The edge placement error of the midpoint evaluation point on each side is calculated based on the fitting results to obtain the weighted edge placement error.

6. The OPC method for parametric curve masks based on piecewise Bézier as described in claim 1, characterized in that, After performing Bézier curve fitting on the new control points and calculating the weighted edge placement error of the evaluation points in the fitting result, the method further includes: Using the new control point as the control point and the weighted edge placement error as the edge placement error, return to the step "Construct a mask error factor matrix based on the evaluation point and control point, with the edge placement error as the target" until the weighted edge placement error of all evaluation points is less than the preset threshold or the number of iterations is reached.

7. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the piecewise Bézier-based parametric curve mask OPC method according to any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the parametric curve mask OPC method based on piecewise Bézier as described in any one of claims 1-6.

9. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the parametric curve mask OPC method based on piecewise Bézier as described in any one of claims 1-6.

Images