Graph extraction optimization method, device and equipment, computer storage medium and program product

Abstract

The application discloses a pattern extraction optimization method and device, equipment, a computer storage medium, and a program product. The method comprises the following steps: acquiring a main pattern in a mask design pattern file; adjusting a sub-resolution auxiliary pattern corresponding to the main pattern to obtain an optimized sub-resolution auxiliary pattern, the optimized sub-resolution auxiliary pattern is used to make the fitting degree between an exposure pattern contour corresponding to the main pattern and a verification pattern contour satisfy a preset fitting degree condition; extracting a ridge point from a pixel point corresponding to the optimized sub-resolution auxiliary pattern, the gradient value of the ridge point is higher than the gradient values of the remaining pixel points in the domain range corresponding to the ridge point; fitting a ridge line by using the ridge point according to a mask rule checking parameter; converting the ridge line into a stepped polyline; and generating a target sub-resolution auxiliary pattern according to the stepped polyline. The method can efficiently and accurately perform pattern extraction on the sub-resolution auxiliary pattern.

Description

Technical Field

[0001] This application belongs to the field of semiconductor technology, and in particular relates to a method, apparatus, device, computer storage medium, and program product for image extraction and optimization. Background Technology

[0002] Sub-Resolution Assist Feature (SRAF) is an optical proximity correction technique widely used in the semiconductor industry to improve the resolution of patterns during integrated circuit manufacturing. SRAF is a small pattern placed near the main pattern, smaller than the resolution of the lithography machine. It scatters light during exposure, helping the main pattern form a more accurate image on the photoresist.

[0003] Traditional solutions primarily involve establishing a set of rules to add corresponding sub-resolution auxiliary graphics around the main graphic. This method is simple to design, computationally inexpensive, and highly manufacturable. However, with continuous advancements in technology, traditional rules struggle to meet high-precision requirements. Especially at advanced process nodes, traditional rectangular patterns are no longer sufficient to meet accuracy demands, while sub-resolution auxiliary graphics with curved structures currently lack adequate hardware support, making them difficult to apply on production lines. Summary of the Invention

[0004] This application provides a method, apparatus, device, computer storage medium, and program product for optimizing image extraction, which can efficiently and accurately extract sub-resolution auxiliary images.

[0005] In a first aspect, embodiments of this application provide a method for optimizing image extraction, the method comprising:

[0006] Retrieve the main graphic from the mask design file;

[0007] The sub-resolution auxiliary graphic corresponding to the main graphic is adjusted to obtain the optimized sub-resolution auxiliary graphic. The optimized sub-resolution auxiliary graphic is used to ensure that the fit between the exposure graphic outline and the verification graphic outline corresponding to the main graphic meets the preset fit condition.

[0008] Ridge points are extracted from the pixels corresponding to the optimized sub-resolution auxiliary graphics. The gradient value of the ridge point is higher than the gradient value of the other pixels in the neighborhood of the ridge point.

[0009] Parameters are checked according to mask rules, and ridge lines are fitted using ridge points;

[0010] Convert the ridge line into a stepped zigzag line;

[0011] A sub-resolution auxiliary graphic of the target is generated based on the stepped polyline. The sub-resolution auxiliary graphic of the target is a Manhattan graphic.

[0012] Secondly, embodiments of this application provide an image extraction and optimization apparatus, the apparatus comprising:

[0013] The acquisition module is used to acquire the main graphic from the mask design drawing file;

[0014] The adjustment module is used to adjust the sub-resolution auxiliary graphic corresponding to the main graphic to obtain the optimized sub-resolution auxiliary graphic. The optimized sub-resolution auxiliary graphic is used to make the fit between the exposure graphic outline and the verification graphic outline corresponding to the main graphic meet the preset fit condition.

[0015] The extraction module is used to extract ridge points from the pixels corresponding to the sub-resolution auxiliary graphics. The gradient value of the ridge point is higher than the gradient value of the other pixels in the neighborhood of the ridge point.

[0016] The fitting module is used to check parameters according to mask rules and fit ridge lines using ridge points;

[0017] A conversion module is used to convert ridge lines into stepped polylines;

[0018] The generation module is used to generate a target sub-resolution auxiliary graphic based on the stepped polyline. The target sub-resolution auxiliary graphic is a Manhattan graphic.

[0019] Thirdly, embodiments of this application provide a graphics extraction and optimization device, the device comprising: a processor and a memory storing computer program instructions; the processor implements the above-described graphics extraction and optimization method when executing the computer program instructions.

[0020] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the above-described graphics extraction optimization method.

[0021] Fifthly, embodiments of this application provide a computer program product, wherein instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform the above-described graphics extraction optimization method.

[0022] In this embodiment, obtaining the main image establishes the foundation for the entire optimization process. Adjusting the sub-resolution auxiliary image corresponding to the main image yields an optimized sub-resolution auxiliary image, ensuring a more precise contour of the main image during exposure. Ridge points are key points in the optimized sub-resolution auxiliary image, typically located at the edges or transitions of the image. Extracting ridge points helps to accurately locate the key features of the optimized sub-resolution auxiliary image, thus providing an accurate data foundation for subsequent ridgeline fitting. Ridge lines are curves connecting ridge points, representing the main contour of the image. By checking parameters according to masking rules and utilizing ridges... Point fitting of ridge lines ensures that the generated ridge lines are highly consistent with the contour of the actual pattern, meeting the requirements of mask rules. Converting ridge lines into stepped polygonal lines simplifies the complexity of the pattern, making it easier to implement in actual photolithography processes. Stepped polygonal lines connect key points with straight line segments, reducing the complexity of curves and improving the accuracy and reliability of photolithography. Based on the stepped polygonal lines, a target sub-resolution auxiliary pattern, namely a Manhattan pattern, is generated. The Manhattan pattern contains only horizontal and vertical line segments, which is easier to implement in photolithography processes, avoiding complex curves and oblique lines, and reducing process difficulty and cost. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic flowchart of the image extraction optimization method provided in the embodiments of this application;

[0025] Figure 2 This is a schematic diagram of an optimized sub-resolution auxiliary graphic provided in an embodiment of this application;

[0026] Figure 3 This is a schematic diagram of a ridge point and ridge line provided in an embodiment of this application;

[0027] Figure 4 This is a schematic diagram of a graphics processing embodiment provided in this application;

[0028] Figure 5 This is a schematic diagram of a Manhattan graphic provided in an embodiment of this application;

[0029] Figure 6 This is a schematic flowchart of a graphic extraction optimization method provided in an embodiment of this application;

[0030] Figure 7 This is a schematic diagram of the structure of the image extraction and optimization device provided in the embodiments of this application;

[0031] Figure 8 This is a schematic diagram of the structure of the graphic extraction and optimization device provided in the embodiments of this application. Detailed Implementation

[0032] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0034] To address the problems of the prior art, embodiments of this application provide a method, apparatus, device, computer storage medium, and computer program product for image extraction and optimization. The image extraction and optimization method provided in this application embodiment will be described first below.

[0035] Figure 1 A schematic flowchart of an embodiment of the image extraction optimization method provided in this application is shown. Figure 1 As shown, Figure 1 The process includes the following steps S101 to S106.

[0036] S101. Obtain the main graphic from the mask design drawing file.

[0037] Mask design files are critical documents in the photolithography process of semiconductor manufacturing. They contain the geometric features and design parameters of the pattern, enabling them to be recognized and used by photolithography equipment. For example, they may include graphic information, which describes the pattern to be etched or deposited on the chip, typically the layout of various circuits; design rules, such as design specifications and constraints regarding linewidth, spacing, and interlayer alignment; and physical properties, such as material selection, thickness, and other parameters that affect the photolithography process.

[0038] Mask design graphics files are the foundation for generating photomasks during the manufacturing process, ensuring that the pattern can be accurately transferred onto the silicon wafer to form the actual circuit structure.

[0039] The master pattern refers to the main graphic outline of interest in the mask design file. It typically represents the basic shape and features of the chip circuitry and usually includes important components such as transistors, interconnects, and resistors. It is primarily represented by two-dimensional geometry. During the photolithography process, the master pattern needs to be accurately transferred onto the silicon wafer to ensure the chip's functionality and performance.

[0040] S102. Adjust the sub-resolution auxiliary graphic corresponding to the main graphic to obtain the optimized sub-resolution auxiliary graphic.

[0041] Optimize the sub-resolution auxiliary graphic to ensure that the fit between the exposure graphic outline corresponding to the main graphic and the verification graphic outline meets the preset fit condition.

[0042] Sub-resolution auxiliary patterns refer to auxiliary patterns used in the photolithography process to support or supplement the main pattern. They are usually located at or near the edge of the main pattern and utilize optical effects to optimize the photolithography process, playing an important role in improving the photolithography quality and resolution of the main pattern.

[0043] In other words, sub-resolution auxiliary patterns are used to improve the imaging quality of the main pattern, reduce optical blur and interference, and help improve the accuracy of pattern transfer. The feature size of the sub-resolution auxiliary pattern is usually smaller than that of the main pattern, and its purpose is to improve the light resolution during photolithography by improving the illumination distribution.

[0044] The exposure pattern outline corresponding to the main pattern refers to the actual optical outline formed on the silicon wafer after exposure during the photolithography process.

[0045] Verify the pattern profile as the pattern profile that is expected to form on the silicon wafer (which can also be understood as the pattern profile set according to design specifications and requirements).

[0046] Fit is an indicator that measures the similarity between the outline of the exposure pattern corresponding to the main pattern and the outline of the verification pattern, and is used to ensure that there is a sufficient matching relationship between the final exposure pattern and the design pattern.

[0047] Preset fit conditions can include evaluation indicators such as line width, spacing, and overall shape of the graphic, which are used to determine whether the outline of the exposed graphic corresponding to the main graphic is close enough to the outline of the verification graphic.

[0048] As an example rather than a limitation, the preset fit conditions can be determined based on the line width tolerance, which specifies that the line width in the main graphic must be within a certain range. For example, assuming the designed line width is 100 nanometers and the allowable tolerance is ±5 nanometers, then the actual line width after exposure should be between 95 nanometers and 105 nanometers.

[0049] As an example, not a limitation, the preset fit conditions can be determined based on the spacing tolerance, that is, defining the minimum distance requirement between logic elements or lines to prevent electrical short circuits or signal interference. For example, assuming that for two adjacent conductors, the spacing is set to 200 nanometers during design and the allowable tolerance is ±10 nanometers, then the actual spacing after exposure must be between 190 nanometers and 210 nanometers.

[0050] As an example rather than a limitation, the preset fit conditions can be determined based on geometric consistency, that is, the requirement for geometric contour consistency for a specific graphic shape (such as a circle, a square, etc.). For example, assuming the design is a circular graphic with a diameter of 50 nanometers, the diameter of the exposed circle should be in the range of 45 nanometers to 55 nanometers to ensure that there is no obvious shape distortion.

[0051] As an example rather than a limitation, the preset fit conditions can be determined based on contour alignment conditions, which require a certain alignment accuracy between the main patterns at different levels (such as multi-layer circuits) to ensure the functionality of the circuit. For example, assuming the alignment requirement between two circuit layers is ±0.5 micrometers, it means that the critical patterns between layers (such as contact holes) need to meet this alignment accuracy to ensure that power and signals can be correctly connected.

[0052] By adjusting the sub-resolution auxiliary pattern, the fit between the exposure pattern outline and the verification pattern outline corresponding to the main pattern can be improved, thereby enhancing the overall effect and accuracy of lithography. This ensures that the exposure pattern of the main pattern and the target verification pattern can achieve the expected functions in the design and meet the requirements of the production process.

[0053] S103. Extract ridge points from the pixels corresponding to the optimized sub-resolution auxiliary graphics.

[0054] The gradient value corresponding to a ridge point is higher than the gradient values ​​of other pixels within the neighborhood of that ridge point.

[0055] As can be understood, a pixel (Picture Element) is the smallest basic unit that makes up a digital image. It is a point on the image that contains color and brightness information, and each image is composed of a large number of pixels.

[0056] Each pixel is typically represented by one or more values, such as 0 and 1 in a black and white image, and RGB values ​​(red, green, and blue channels) in a color image. In a two-dimensional image, the coordinates of a pixel are usually represented by its row and column, for example, (10, 15) represents the pixel in the 10th row and 15th column.

[0057] Ridge points are feature points extracted from image analysis. They typically represent points in an image that exhibit significant changes or features, and can also be understood as points used to represent local extrema. In this application, ridge points are the locations with the largest gradient values ​​in a certain feature (such as an edge), and often refer to important features of a graphic or contour.

[0058] Gradient values ​​are a measure of the rate of change of pixel values ​​in an image, representing the degree of change in light intensity in the image. They can be used to detect edges, contours, and features.

[0059] In a two-dimensional image, the gradient is typically represented by a vector, with one component representing the horizontal change (x-direction) and the other representing the vertical change (y-direction). Commonly used calculation methods include the Sobel operator and the Prewitt operator. High gradient values ​​usually indicate the presence of edges or significant features in the image, because these areas exhibit large pixel variations.

[0060] This can also be understood as the gradient value measuring the rate of change of pixel values, used to detect areas of pixel intensity change, especially at edges. A ridge's gradient value is higher than the gradient values ​​of other pixels in its neighborhood, meaning the ridge is more significant in the image and is an important feature location.

[0061] In the field of digital image processing, a neighborhood is a set of other pixels that are adjacent to a particular pixel. It can also be understood as a neighborhood used to describe the characteristics of a local area around a ridge point.

[0062] The center pixel in the neighborhood is a ridge point. The neighborhood can be a region of different shapes and sizes, such as a rectangular neighborhood (e.g., a 3*3 rectangle), a circular neighborhood (a circular region with a radius of r centered at a certain point), a custom shape (set according to a specific application scenario), etc. The size range of the center of the neighborhood is not limited here.

[0063] Ridge points are key points that characterize the local features of a graphic. Extracting ridge points from the pixels corresponding to the optimized sub-resolution auxiliary graphic can effectively help with subsequent ridge line fitting and shape analysis.

[0064] S104. Check the parameters according to the mask rules and fit the ridge line using the ridge point.

[0065] It can be understood that mask rules refer to the rules that ensure the pattern structure meets specific design and manufacturing requirements during the design of circuit patterns (such as mask patterns).

[0066] Mask Rule Check (MRC) parameters are specific metrics or conditions used to verify whether design rules are followed, such as minimum linewidth, minimum spacing, graphic alignment requirements, and coverage area. These parameters use algorithms to check design compliance, ensuring the circuit functions correctly in actual electrical manufacturing.

[0067] A ridge is a line formed by connecting a series of ridge points. It is often used to represent important features or outlines in an image or geometric shape, such as describing the main direction or structure of a shape.

[0068] As can be understood, fitting a ridge line using ridge points refers to using the obtained ridge points (points with obvious features in the image) to generate a fitted line (ridge line) that can optimally pass through or approximate these ridge points.

[0069] As an example rather than a limitation, various mathematical curve fitting algorithms (such as polynomial fitting, spline fitting, etc.) can be used to ensure that the resulting ridges visually match the distribution of ridge points.

[0070] By fitting ridge points to obtain ridge lines, subsequent analysis can be simplified, facilitating further shape analysis, feature extraction, and pattern matching.

[0071] S105. Convert the ridge line into a stepped broken line.

[0072] A stepped line is a broken line composed of a series of line segments. The stepped line has abrupt changes at each turning point, forming a stepped effect. For example, there are significant changes in any two mutually perpendicular directions.

[0073] A stepped zigzag line is a ridge line formed by a series of stepped zigzag lines.

[0074] As an example and not a limitation, there are several methods to convert ridges into stepped broken lines, including but not limited to the following:

[0075] 1. Discretization and quantization

[0076] The points on the ridge line are sampled at fixed increments (such as uniform intervals) to obtain discrete points.

[0077] A stepped polyline is formed by quantizing the coordinates of discrete points to the nearest step level.

[0078] 2. Step-by-step algorithm

[0079] Using a preset algorithm (such as a maximum-minimum strategy), the ridge line is sampled, and a stepped polyline is constructed based on the changes in these sampling points. For example, a simple threshold can be used to replace the Y value with a fixed value in a certain segment to form a step.

[0080] 3. Piecewise linear fitting

[0081] Represent the ridge line using a piecewise linear function, select nodes, and ensure that there is a horizontal or vertical line segment between every two nodes. By selecting parameters, control the coordinates of each node on the polyline, so that a stepped structure is formed between the nodes.

[0082] 4. Image processing tools

[0083] After contour extraction using an image processing library, the contours are simplified and segmented to generate a stepped polyline. This stepped polyline simplifies the representation of complex shapes and highlights their key features.

[0084] It is understandable that stepped polylines reduce the complexity of image shapes, making the analysis, comparison, and processing of image shapes more efficient and simpler. For complex images or geometric shapes, simplified representations facilitate subsequent processing and storage, and stepped polylines can highlight the main features of the image and reduce interference.

[0085] S106. Generate sub-resolution auxiliary graphics of the target based on the stepped polyline.

[0086] The target sub-resolution auxiliary graphic is a Manhattan graphic.

[0087] By using the stepped polyline segments generated, corresponding auxiliary graphics can be created by drawing the segments in a Manhattan structure (i.e., composed of horizontal and vertical line segments).

[0088] For example, based on the stepped polyline, key inflection points are identified, and these inflection points are connected to form a Manhattan shape, making the graphic adaptable to the needs of sub-resolution construction.

[0089] Manhattan graphics are a specific type of design where the shape typically appears as a pattern composed of horizontal and vertical line segments. This form of graphic is regular and can effectively interact with the diffraction properties of light systems.

[0090] When the beam passes through a sub-resolution auxiliary pattern with a Manhattan shape, diffraction and interference enable good accuracy in the final image, even though the actual feature details of the auxiliary pattern cannot be directly observed at conventionally defined resolution.

[0091] The process of generating sub-resolution assisted Manhattan graphics through stepped polylines combines simplified shape representation and optical processing principles, enabling the realization of complex patterns with high precision in actual manufacturing.

[0092] This invention, through obtaining the main image, establishes the foundation for the entire optimization process. Adjusting the corresponding sub-resolution auxiliary image yields an optimized sub-resolution auxiliary image, ensuring a more precise contour of the main image during exposure. Ridge points are key points in the optimized sub-resolution auxiliary image, typically located at the edges or transitions of the image. Extracting ridge points helps accurately locate key features of the optimized sub-resolution auxiliary image, providing an accurate data foundation for subsequent ridgeline fitting. Ridge lines are curves connecting ridge points, representing the main contour of the image. By checking parameters according to masking rules and fitting ridge lines using ridge points, the generated ridge lines can be ensured. The contours of the actual pattern are highly consistent with the mask rules. Converting the ridge lines into stepped polygonal lines simplifies the pattern's complexity, making it easier to implement in actual photolithography processes. The stepped polygonal lines connect key points with straight line segments, reducing the complexity of curves and improving the accuracy and reliability of photolithography. Based on the stepped polygonal lines, a target sub-resolution auxiliary pattern, namely the Manhattan pattern, is generated. The Manhattan pattern contains only horizontal and vertical line segments. This type of pattern is easier to implement in photolithography processes because it avoids complex curves and oblique lines, ensuring optimal photolithography results in actual production while reducing process difficulty and cost.

[0093] In one implementation, the sub-resolution auxiliary graphic corresponding to the main graphic is adjusted to obtain an optimized sub-resolution auxiliary graphic. This includes: setting sub-resolution auxiliary graphic optimization evaluation points corresponding to the boundary of the main graphic, wherein the distance between the sub-resolution auxiliary graphic optimization evaluation points and the boundary of the main graphic meets a preset distance condition; inputting the main graphic and the sub-resolution auxiliary graphic optimization evaluation points into a machine learning model for simulation calculation to obtain the exposure graphic contour of the main graphic, the simulation calculation is used to simulate the photolithography process, adjusting the intensity values ​​corresponding to the sub-resolution auxiliary graphic optimization evaluation points until the fit between the exposure graphic contour and the verification graphic contour corresponding to the main graphic meets a preset fit condition; and adjusting the shape of the sub-resolution auxiliary graphic according to the intensity values ​​corresponding to the sub-resolution auxiliary graphic optimization evaluation points to obtain the optimized sub-resolution auxiliary graphic.

[0094] Sub-resolution auxiliary pattern optimization evaluation points (SRAF points) are specific measurement points set near the boundary of the main pattern during the photolithography process. These points are used to evaluate the relationship between the main pattern and its corresponding sub-resolution auxiliary pattern and its optimization effect.

[0095] Sub-resolution auxiliary image optimization evaluation points are usually selected at the edges or features of the main image. These locations have a significant impact on the final imaging effect and therefore require special attention.

[0096] Sub-resolution auxiliary pattern optimization evaluation points can guide the adjustment process of sub-resolution auxiliary patterns. Based on the feedback obtained from these points, designers can optimize parameters such as the shape and intensity of the auxiliary patterns to improve the imaging quality during the lithography process.

[0097] It is understandable that the sub-resolution auxiliary graphics optimization evaluation points and the boundary of the main graphics usually need to meet a certain preset distance condition to ensure that they can effectively represent the features of the main graphics, and that the evaluation results will not be distorted due to the distance being too close or too far.

[0098] The preset distance condition is a distance limit set between the sub-resolution auxiliary graphic optimization evaluation point and the boundary of the main graphic to ensure the effectiveness and accuracy of the optimization evaluation.

[0099] As an example rather than a limitation, preset distance conditions may include minimum distance, maximum distance, angle limit, relative position constraint, etc.

[0100] Minimum distance refers to the minimum allowable distance between the sub-resolution auxiliary pattern optimization evaluation point and the boundary of the main pattern. This condition ensures that the evaluation point is not too close to the boundary, thereby avoiding data distortion or imaging noise due to edge effects, preventing unnecessary interference during the photolithography process, and ensuring that the evaluation point can accurately reflect the characteristics of the main pattern.

[0101] Maximum distance refers to the maximum distance limit between the sub-resolution auxiliary image evaluation point and the boundary of the main image. This condition prevents the evaluation point from being too far from the boundary, which would prevent it from effectively capturing the influence of the main image on the imaging, and ensures that the optimized evaluation point can effectively represent the optical characteristics of the main image.

[0102] Angle constraints refer to the fact that, in some cases, the positions of sub-resolution auxiliary graphic evaluation points may also need to follow specific angle constraints (e.g., the distance relationship between the evaluation point and the boundary of the main graphic in a specific direction). This helps to maintain the geometric continuity and stability of the main graphic, especially in complex shapes or curved sections.

[0103] Relative position constraints refer to considering the relative positions of multiple evaluation points. For example, within a specific area, multiple evaluation points should be evenly distributed or arranged in a specific manner. This ensures that rich feedback information is obtained during the evaluation process, and more comprehensively reflects the characteristics of the main graphic.

[0104] It is understandable that the data corresponding to the optimization evaluation points of the main image and each sub-resolution auxiliary image are input into the machine learning model, and simulation calculations are performed using methods such as backpropagation and optimization algorithms to simulate the behavior of the interaction between the light beam and the image during the photolithography process.

[0105] As an example, and not a limitation, the machine learning models mentioned here can include: regression models, such as linear regression, ridge regression, Lasso regression, etc., to predict the relationship between light intensity and location; support vector machines (SVMs), which are used for classification or regression to predict exposure pattern contours under different conditions; neural networks, such as deep learning models like convolutional neural networks (CNNs), which can handle more complex nonlinear relationships; generative adversarial networks (GANs), used to generate high-quality exposure pattern contours. Through adversarial training, more realistic images can be simulated, which can then be used to verify and optimize the lithography process; variational autoencoders (VAEs), which can learn the latent distribution of data, generate new data points, and help explore the lithography effects under different behavioral conditions.

[0106] It is understandable that different machine learning models may be used in different manufacturing processes, and the specific machine learning model can be selected according to the actual situation.

[0107] In other words, through simulation calculations, the actual photolithography process can be simulated to obtain the exposure pattern outline of the main pattern, thereby evaluating the performance of the original design and sub-resolution auxiliary pattern in order to detect potential development defects.

[0108] Intensity value can be understood as the light intensity of light emitted by a light source at a specific point (such as the main graphic or optimization evaluation point).

[0109] During photolithography, the light intensity at different locations varies due to factors such as the intensity of the light source, the light propagation path, refraction, and reflection. Light intensity directly affects the chemical reaction of the photoresist. For example, excessively high light intensity may lead to overexposure, thus affecting the sharpness and shape of the pattern edges; while insufficient light intensity may result in incomplete development of the pattern.

[0110] The fit between the exposure pattern contour and the verification pattern contour in the simulation results is evaluated, and the intensity value corresponding to each sub-resolution auxiliary pattern optimization evaluation point is adjusted as needed. For example, gradient descent or other optimization algorithms can be used to gradually improve the fit.

[0111] By adjusting the intensity values ​​corresponding to the evaluation points of the sub-resolution auxiliary pattern, the sub-resolution auxiliary pattern can be further optimized to produce a more suitable light intensity distribution during the photolithography process, ensuring that the simulated exposure effect is as close as possible to the expectation. After optimization, the photolithography efficiency can be improved.

[0112] The optimized intensity value is mapped onto the shape features of the sub-resolution auxiliary pattern. By adjusting the width, spacing, or other geometric parameters of the shape, such as by using geometric transformation and line segment weighting, the shape of the sub-resolution auxiliary pattern can be optimized. The optimized sub-resolution auxiliary pattern can be obtained, which can be updated relative to the main pattern during the lithography process, so that the actual imaging effect is closer to the design requirements, and the lithography accuracy and yield are improved.

[0113] Figure 2 This is a schematic diagram of an optimized sub-resolution auxiliary graphic provided in an embodiment of this application.

[0114] like Figure 2 As shown, the shape of the sub-resolution auxiliary graphic is adjusted by adjusting the intensity value corresponding to the evaluation point of the sub-resolution auxiliary graphic optimization, and finally the optimized sub-resolution auxiliary graphic is obtained.

[0115] After obtaining the optimized sub-resolution auxiliary graphic, further graphic extraction can be performed on the optimized sub-resolution auxiliary graphic itself.

[0116] In one implementation, the parameters are checked according to the mask rules, and the ridge line is fitted using ridge points. This includes: checking the parameters according to the mask rules, grouping two ridge points with adjacent spatial coordinates into a group to obtain multiple ridge point groups; fitting the ridge points in the same ridge point group into fitted line segments; traversing all ridge point groups and connecting the fitted line segments in sequence to obtain the ridge line.

[0117] As an example rather than a limitation, the MRC parameters include various parameters such as minimum line width, minimum spacing, graphic alignment requirements, and coverage area. Based on these predefined parameters, the distance between adjacent ridge points is checked to see if it is less than a distance threshold parameter. If the distance between two ridge points is less than a specific threshold, they are grouped together.

[0118] Suppose there is a set of ridge points with the following coordinates: P1(1,2), P2(1,3), P3(5,5), P4(6,6). According to the masking rule, if the maximum distance between the ridge points is 2, then P1 and P2 can form one group, while P3 and P4 can form another group.

[0119] Grouping ridge points using masking rules can simplify the subsequent fitting process and improve efficiency and accuracy.

[0120] Perform linear regression or other types of line fitting on all ridge points within the same group to obtain the best-fit line segment. For example, use the least squares method to minimize the distance between the fitted line segment and the actual ridge point.

[0121] Suppose that for the ridge point set {P1(1,2), P2(1,3)}, a line segment L1 can be chosen to fit these points. The expression corresponding to the line segment L1 is obtained by using the least squares method.

[0122] Iterate through all fitted line segments, find the connection points between these segments, and connect them to form a complete ridge. For example, if the end point of the current line segment is close to the start point of the next line segment, they can be considered adjacent and connected. In the example above, the rule is to connect adjacent ridge point groups. Assume we have fitted line segments with the following ridge point groups:

[0123] Line segment 1: L1 (connecting P1 and P2);

[0124] Line segment 2: L2 (connecting P3 and P4).

[0125] Fitting line segments provides a simplified way of representing the overall trend of a ridge point group, which helps reduce the complexity of subsequent processing.

[0126] It is understandable that when connecting these line segments, if the endpoint of L1 meets the starting point of L2, a continuous ridge line can be formed. This method can transform the original scattered ridge points into smooth ridge lines, improving the efficiency of data processing.

[0127] Figure 3 This is a schematic diagram of a ridge point and ridge line provided in an embodiment of this application.

[0128] Combination Figure 2 and Figure 3 It is obvious that... Figure 2 Ridge point extraction is performed using optimized sub-resolution auxiliary graphics, and line fitting is performed by traversing all ridge points to obtain the desired result. Figure 3 A schematic diagram of the graphic extraction formed by ridge points and ridge lines.

[0129] In one implementation, converting a ridge line into a stepped polyline includes: determining an interval value based on masking rules, where the interval value represents the distance between any two adjacent ridge points; extracting key points from the ridge points based on the interval value; and connecting any adjacent key points with a polyline to convert the ridge line from a fitted line into a stepped polyline.

[0130] It is understandable that the mask rule check parameters can define the distance that should be maintained between any two adjacent ridge points, and determine the appropriate interval value to filter ridge points to ensure that they comply with manufacturing process and physical constraints.

[0131] Suppose we have a set of ridge points for a ridge line:

[0132] Ridge coordinates: P1(1,1), P2(2,3), P3(2,5), P4(5,7), P5(7,8).

[0133] Assuming the set interval value is 2, this means that the search for the next ridge point starts from the initial ridge point and continues until this interval is reached.

[0134] As an example rather than a limitation, starting from any ridge point, you can filter the next ridge point by a certain number of points or by a certain distance, thus obtaining the key points.

[0135] Connect the extracted key points with polylines, traverse all key points, and connect adjacent key points in turn to form a stepped polyline.

[0136] This process simplifies the original ridge line to key points containing the main structural changes, making subsequent processing more focused and efficient.

[0137] In one implementation, connecting any adjacent key points with a polyline to convert the ridge line from a fitted line into a stepped polyline includes: extending a first straight line from any key point until the first straight line intersects with a second straight line extended from a key point adjacent to that key point to obtain a turning point; connecting the key points and the turning points in sequence to obtain a stepped polyline, wherein the first straight line and the second straight line are perpendicular to each other.

[0138] Choose any key point from the extracted key point set as the starting point, and extend a straight line (the first straight line) in either the horizontal or vertical direction until this straight line intersects with the straight line extended from the next key point adjacent to its starting point (the second straight line). The first straight line and the second straight line are perpendicular, and the intersection of the two straight lines is the turning point.

[0139] Suppose the following key points exist: K1(1,1), K2(2,4), K3(5,7).

[0140] Starting from K1(1,1), extend to K2, where a straight line is extended horizontally, and then a straight line is extended vertically from K2(2,4). The intersection of these two lines is the turning point (2,1).

[0141] The stepped broken line obtained by connecting K1, the turning point, and K2 is: K1(1,1)→(2,1)→K2(2,4), where the point (2,1) is the turning point.

[0142] Select key points in the extracted key point list in sequence, and repeat the above extension process for each key point to connect all key points and all turning points into a whole stepped polyline.

[0143] In other words, starting from K2(2,4), extend a straight line horizontally, and from K3(5,7), extend a straight line vertically. These two lines intersect at point (5,4), forming a turning point. Connecting K2, the turning point, and K3, we obtain a stepped broken line: K2(2,4)→(5,4)→K3(5,7). This process continues until the overall stepped broken line is obtained.

[0144] By forming a stepped polyline, the representation of ridges is simplified, making the geometric characteristics of the graph more apparent and facilitating subsequent analysis. The polyline form is more efficient in computation and optimization, reducing computational resource consumption compared to the original unordered ridges.

[0145] In one implementation, generating a target sub-resolution auxiliary graphic based on a stepped polyline includes: extracting inflection points from the stepped polyline, where the inflection points are the endpoints of the polyline segments; recording the coordinates of the inflection points to obtain step points; obtaining rectangle parameters from mask rule check parameters, where the rectangle parameters are used to determine the side length of the rectangle; calculating the spatial coordinates of the rectangle points corresponding to the two adjacent step points based on the spatial coordinates of the two adjacent step points and the rectangle parameters to obtain rectangle points; starting from any rectangle point, sequentially connecting the rectangle points with straight lines to generate rectangles corresponding to the two adjacent step points; traversing all step points to generate rectangles corresponding to all step points; starting from any rectangle, traversing all rectangles, sequentially connecting the outer contours of the entire area covered by all rectangles to generate a Manhattan graphic, and determining the Manhattan graphic as the target sub-resolution auxiliary graphic.

[0146] A turning point is the endpoint of each segment in a stepped polyline. It can also be understood that each turning point is usually a key point or the endpoint of a polyline segment connecting two adjacent key points.

[0147] Suppose the inflection points of the stepped polyline we extracted are: P1(1,1), P2(2,1), P3(4,3), P4(4,5), P5(6,5). Record the coordinates of these points to obtain the set of stepped points.

[0148] The rectangle's side length is determined by checking parameters according to mask rules. This can be a fixed value or defined by the user.

[0149] Suppose the obtained rectangle parameters are as follows: rectangle length (L) = 3 units; rectangle width (H) = 2 units.

[0150] Taking P1 and P2 as an example, calculate the coordinates (x, y) of the midpoint between P1 and P2, that is, (1+2) / 2=1.5, (1+1) / 2=1, and get the coordinates of the midpoint as (1.5,1).

[0151] Using the midpoint as the center point of the rectangle, and combining the rectangle's length and width, calculate the four vertices of the rectangle. According to relevant mathematical principles,

[0152] The coordinates of the vertex located at the top left corner of the center point are (xL / 2, y+H / 2), which can be obtained by substituting into the example above (0,2).

[0153] The coordinates of the vertex located at the upper right corner of the center point are (x+L / 2, y+H / 2), which can be obtained by substituting into the example above (3,2).

[0154] The coordinates of the vertex located at the lower left corner of the center point are (xL / 2, yH / 2), which can be obtained by substituting into the example above (0,0).

[0155] The coordinates of the vertex located at the bottom right corner of the center point are (x+L / 2, yH / 2), which can be obtained by substituting into the example above (3,0).

[0156] In other words, the rectangle generated based on points P1 and P2 and the rectangle parameters is the rectangle formed by connecting the four points (0,0), (3,0), (3,2), and (0,2).

[0157] Starting from any given rectangular point, connect the four calculated rectangular points sequentially to form a rectangle. Repeat the above steps for generating rectangles for each pair of adjacent points to form a series of rectangles. Traverse all generated rectangles and connect their outer contours sequentially to form the global Manhattan graphic.

[0158] Figure 4 This is a schematic diagram of a graphics processing method provided in an embodiment of this application.

[0159] like Figure 4 As shown, firstly, keypoints (red points) are determined from the ridge points (yellow points) based on the interval values. Then, each keypoint is connected sequentially using a polyline. Next, using the keypoints as a reference, the step points at the turning points are determined. Connecting all turning points sequentially with straight lines transforms the basic polyline into a step-shaped polyline. Then, rectangles corresponding to two adjacent step points are generated based on their coordinates and corresponding rectangle parameters. This process is repeated for all step points, generating rectangles for all step points. Connecting all rectangles sequentially to cover the entire outer contour of the area yields the Manhattan graphic corresponding to the optimized sub-resolution auxiliary graphic.

[0160] Figure 5 This is a schematic diagram of a Manhattan graphic provided in an embodiment of this application.

[0161] like Figure 5As shown, the Manhattan-shaped sub-resolution auxiliary pattern can effectively improve the pattern resolution during the photolithography process, helping to achieve smaller feature sizes, thereby improving chip integration and performance. Through optimized design, the Manhattan-shaped sub-resolution auxiliary pattern can better improve the uniformity of the photolithography pattern, reduce the defect rate, and thus improve production yield. The Manhattan-shaped sub-resolution auxiliary pattern can be flexibly extracted and adjusted according to the mask manufacturing hardware conditions, making it suitable for production needs under various conditions.

[0162] Figure 6 This is a flowchart illustrating a graphic extraction optimization method provided in an embodiment of this application.

[0163] like Figure 6 As shown, Figure 6 The process includes the following steps S601 to S613.

[0164] S601. Input the GDS file.

[0165] In conjunction with the above, this refers to inputting the mask design graphic file.

[0166] S602. Place sub-resolution auxiliary graphic evaluation points around the main graphic.

[0167] Sub-resolution auxiliary image optimization evaluation points are usually selected at the edges or features of the main image. These locations have a significant impact on the final imaging effect and therefore require special attention.

[0168] S603. Input the main graphic and evaluation points into the model for collaborative optimization.

[0169] Subresolution auxiliary image optimization evaluation points (SRAF points) are placed around the main image and input into the model along with the main image for optimization. The intensity values ​​of the SRAF points are adjusted based on the simulation results, and then simulation calculations are performed again. This process is repeated until the simulated exposure image contour sufficiently matches the verification image contour, resulting in the optimized subresolution auxiliary image (e.g., ...). Figure 2 ).

[0170] S604, pixelation processing.

[0171] The optimized sub-resolution auxiliary graphics are pixelated to generate the original matrix.

[0172] S605, Calculate the gradient value.

[0173] Calculate the gradient value of each pixel in the original matrix to obtain the gradient value image.

[0174] S606, Mark the ridge point.

[0175] Combining the above text Figure 3Local peak points are extracted from the gradient value image, their coordinates are recorded and marked as ridge points.

[0176] S607. Group the ridge points according to the MRC parameters.

[0177] The parameters are checked according to the mask rules, and adjacent ridge points are extracted and grouped.

[0178] S608, Extract the ridge line.

[0179] A curve is fitted to the ridge points of the same group and called a ridgeline.

[0180] S609. Extract key points based on MRC parameters.

[0181] Based on the MRC parameters, points are taken at intervals along the ridge line, and their coordinates are recorded as key points.

[0182] S610. Convert the ridge line into a stepped broken line based on the key points.

[0183] Based on key points, the ridge lines are converted into stepped broken lines.

[0184] S611, Record the step points.

[0185] Record the coordinates of the turning point and mark it as a step point.

[0186] S612. Based on the MRC parameters, generate a rectangular pattern with the ends connected by the step points.

[0187] Based on the MRC parameters, determine the rectangle parameters, and generate a rectangle based on the rectangle parameters and the coordinates between two adjacent step points.

[0188] S613, Obtain the target sub-resolution auxiliary graphics.

[0189] Combining the above text Figure 5 The Manhattan shape formed by connecting all the rectangles is the final goal.

[0190] In summary, using Manhattan-shaped subresolution auxiliary patterns can provide higher pattern resolution, pattern uniformity, and greater flexibility to adapt to production conditions, ultimately improving the quality and efficiency of mask optimization. In the highly competitive semiconductor industry, these advantages make Manhattan-shaped subresolution auxiliary patterns an important design and manufacturing tool. Using Manhattan patterns as target subresolution auxiliary patterns can be used to improve resolution and manufacturability during photolithography.

[0191] Based on the image extraction optimization method provided in the above embodiments, this application also provides specific implementation methods of the image extraction optimization device. Please refer to the following embodiments.

[0192] Figure 7 This is a schematic diagram of the structure of the graphic extraction and optimization device provided in the embodiments of this application.

[0193] First see Figure 7 The image extraction and optimization device provided in this application embodiment includes the following modules:

[0194] Module 1001 is used to acquire the main graphic from the mask design drawing file;

[0195] The adjustment module 1002 is used to adjust the sub-resolution auxiliary graphic corresponding to the main graphic to obtain an optimized sub-resolution auxiliary graphic. The optimized sub-resolution auxiliary graphic is used to make the fit between the exposure graphic outline and the verification graphic outline corresponding to the main graphic meet the preset fit condition.

[0196] The extraction module 1003 is used to extract ridge points from the pixels corresponding to the optimized sub-resolution auxiliary graphics. The gradient value of the ridge point is higher than the gradient value of the other pixels in the neighborhood of the ridge point.

[0197] Fitting module 1004 is used to check parameters according to mask rules and fit ridge lines using ridge points;

[0198] Conversion module 1005 is used to convert ridge lines into stepped polylines;

[0199] The generation module 1006 is used to generate a target sub-resolution auxiliary graphic based on the stepped polyline. The target sub-resolution auxiliary graphic is a Manhattan graphic.

[0200] Obtaining the main image establishes the foundation for the entire optimization process. Adjusting the corresponding sub-resolution auxiliary image yields an optimized sub-resolution auxiliary image, ensuring a more precise contour of the main image during exposure. Ridge points are key points in the optimized sub-resolution auxiliary image, typically located at edges or transitions. Extracting ridge points helps accurately locate key features of the optimized sub-resolution auxiliary image, providing an accurate data foundation for subsequent ridgeline fitting. Ridges are curves connecting ridge points, representing the main contour of the image. By checking parameters according to masking rules and fitting ridge lines using ridge points, it ensures that the generated ridge lines match the actual contours. The contours of the patterns are highly consistent, conforming to the requirements of mask rules. Converting ridge lines into stepped polygonal lines simplifies the complexity of the patterns, making them easier to implement in actual photolithography processes. The stepped polygonal lines connect key points through straight line segments, reducing the complexity of curves and improving the accuracy and reliability of photolithography. Based on the stepped polygonal lines, a target sub-resolution auxiliary pattern, namely the Manhattan pattern, is generated. The Manhattan pattern contains only horizontal and vertical line segments. This type of pattern is easier to implement in photolithography processes because it avoids complex curves and oblique lines, ensuring optimal photolithography results in actual production while reducing process difficulty and cost.

[0201] As one implementation of this application, the adjustment module is further used to adjust the sub-resolution auxiliary graphic corresponding to the main graphic to obtain an optimized sub-resolution auxiliary graphic, including: setting sub-resolution auxiliary graphic optimization evaluation points corresponding to the boundary of the main graphic, wherein the distance between the sub-resolution auxiliary graphic optimization evaluation points and the boundary of the main graphic meets a preset distance condition; inputting the main graphic and the sub-resolution auxiliary graphic optimization evaluation points into a machine learning model for simulation calculation to obtain the exposure graphic contour of the main graphic, wherein the simulation calculation is used to simulate the photolithography process; adjusting the intensity value corresponding to the sub-resolution auxiliary graphic optimization evaluation points until the fit between the exposure graphic contour and the verification graphic contour corresponding to the main graphic meets a preset fit condition; and adjusting the shape of the sub-resolution auxiliary graphic according to the intensity value corresponding to the sub-resolution auxiliary graphic optimization evaluation points to obtain an optimized sub-resolution auxiliary graphic.

[0202] Setting sub-resolution auxiliary pattern optimization evaluation points corresponding to the boundaries of the main pattern can determine key positions near the boundaries of the main pattern. The main pattern and sub-resolution auxiliary pattern optimization evaluation points are input into a machine learning model for simulation calculations. The machine learning model simulates the lithography process and predicts the exposure pattern contour of the main pattern under different sub-resolution auxiliary pattern configurations. This helps to quickly evaluate the effects of different configurations and find the optimal sub-resolution auxiliary pattern layout. By adjusting the intensity values ​​of the optimization evaluation points, the configuration of the sub-resolution auxiliary patterns is gradually optimized, ensuring that the fit between the exposure pattern contour of the main pattern and the verification pattern contour meets preset conditions. This step ensures that the final generated pattern achieves the expected accuracy and effect in the actual lithography process. Based on the simulation calculation results, the shape of the sub-resolution auxiliary patterns is finely adjusted to finally generate optimized sub-resolution auxiliary patterns. Through simulation calculations and machine learning models, the optimal sub-resolution auxiliary pattern configuration can be quickly found, reducing trial and error and improving the manufacturability and production efficiency of the process.

[0203] As one implementation of this application, the fitting module is also used to check parameters according to mask rules and fit ridge lines using ridge points, including: checking parameters according to mask rules, grouping two ridge points with adjacent spatial coordinate positions into a group to obtain multiple ridge point groups; fitting ridge points in the same ridge point group into fitting line segments; traversing all ridge point groups and connecting the fitting line segments in sequence to obtain ridge lines.

[0204] By checking parameters according to masking rules, adjacent ridge points are grouped together to ensure that the generated ridge lines meet the hardware requirements for mask manufacturing. Fitting ridge points within the same group to fitted line segments simplifies complex graphic representations, reduces data volume and computational complexity, and makes fitted line segments easier to process and optimize, thus improving computational efficiency. By traversing all ridge point groups and sequentially connecting the fitted line segments, a continuous ridge line is obtained. This ridge line accurately represents the main outline of the sub-resolution auxiliary graphic, providing a foundation for subsequent graphic generation and optimization.

[0205] As one implementation of this application, the conversion module is also used to convert the ridge line into a stepped polyline, including: determining the interval value according to the mask rule check parameters, the interval value being used to represent the interval distance between any two adjacent ridge points; extracting key points from the ridge points according to the interval value; and connecting any adjacent key points with a polyline to convert the ridge line from a fitted line into a stepped polyline.

[0206] Using mask rules to guide polyline generation avoids manufacturing problems caused by hardware incompatibility, improving manufacturability and reliability. Converting ridges into stepped polylines simplifies graphic representation, reduces computational complexity and data processing, accelerates graphic generation and optimization, and improves overall computational efficiency. Extracting key points and connecting them with polylines allows for more precise control over graphic details, ensuring high precision and quality during photolithography. Stepped polylines provide a foundation for sub-resolution assisted graphic layout optimization, making the optimization process more intuitive and efficient, and contributing to higher graphic resolution and shape capture accuracy.

[0207] As one implementation of this application, the fitting module is also used to connect any adjacent key points with a polyline and convert the ridge line from the fitting line into a stepped polyline, including: extending a first straight line from any key point until the first straight line intersects with a second straight line extended from a key point adjacent to the key point to obtain a turning point, and connecting the key points and the turning point in sequence to obtain a stepped polyline, wherein the first straight line and the second straight line are perpendicular to each other.

[0208] By converting ridges into stepped polygonal lines, the generated polygons meet the specific requirements of mask manufacturing, reducing errors in the manufacturing process. Connecting key points with polygonal lines transforms the ridges from fitted lines into stepped polygons, further simplifying the graphical representation, reducing data volume and computational complexity, and improving computational efficiency. Stepped polygons are easier to implement in actual manufacturing, especially in semiconductor manufacturing with high precision requirements. This form of ridge reduces errors in the manufacturing process and improves the reliability and consistency of the process.

[0209] As one implementation of this application, the generation module is further configured to generate a target sub-resolution auxiliary graphic based on a stepped polyline, including: extracting the turning points in the stepped polyline, where the turning points are the endpoints of the polyline segments; recording the coordinates of the turning points to obtain the step points; obtaining rectangle parameters from the mask rule check parameters, where the rectangle parameters are used to determine the side length of the rectangle; calculating the spatial coordinates of the rectangle points corresponding to the two adjacent step points based on the spatial coordinates of the two adjacent step points and the rectangle parameters to obtain the rectangle points; starting from any rectangle point, sequentially connecting the rectangle points with straight lines to generate the rectangles corresponding to the two adjacent step points; traversing all step points to generate rectangles corresponding to all step points; starting from any rectangle, traversing all rectangles, sequentially connecting the outer contours of the entire area covered by all rectangles to generate a Manhattan graphic, and determining the Manhattan graphic as the target sub-resolution auxiliary graphic.

[0210] By extracting the inflection points of the stepped polyline and recording their coordinates, stepped points are generated, ensuring that the generated sub-resolution auxiliary pattern meets the specific requirements of mask manufacturing. Calculating the spatial coordinates of rectangular points based on the spatial coordinates of two adjacent stepped points and the rectangle parameters allows for more precise control over the accuracy of the generated rectangles. By generating rectangles and connecting their points, a Manhattan pattern is created, further simplifying the graphic representation. Manhattan patterns are easier to implement in actual manufacturing, especially in semiconductor manufacturing with high precision requirements. This form of pattern reduces errors in the manufacturing process and improves the reliability and consistency of the process.

[0211] Figure 8 This is a schematic diagram of the structure of the graphic extraction and optimization device provided in the embodiments of this application.

[0212] The graphics extraction and optimization device may include a processor 2001 and a memory 2002 storing computer program instructions.

[0213] Specifically, the processor 2001 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0214] Memory 2002 may include mass storage for data or instructions. For example, and not limitingly, memory 2002 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 2002 may include removable or non-removable (or fixed) media. Where appropriate, memory 2002 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 2002 is non-volatile solid-state memory.

[0215] In certain embodiments, the memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Thus, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the methods according to one aspect of this disclosure.

[0216] The processor 2001 reads and executes computer program instructions stored in the memory 2002 to implement any of the graphics extraction optimization methods in the above embodiments.

[0217] In one example, the image extraction and optimization device may further include a communication interface 2003 and a bus 2000. For example, Figure 8 As shown, the processor 2001, memory 2002, and communication interface 2003 are connected through bus 2000 and complete communication with each other.

[0218] The communication interface 2003 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0219] Bus 2000 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 2000 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0220] Furthermore, in conjunction with the image extraction optimization methods described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the image extraction optimization methods described in the above embodiments.

[0221] This application also provides a computer program product, including a computer program, which, when executed, implements any of the graphic extraction optimization methods described in the above embodiments.

[0222] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0223] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0224] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0225] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0226] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for optimizing image extraction, characterized in that, The method includes: Retrieve the main graphic from the mask design file; The sub-resolution auxiliary graphic corresponding to the main graphic is adjusted to obtain an optimized sub-resolution auxiliary graphic. The optimized sub-resolution auxiliary graphic is used to make the fit between the exposure graphic outline and the verification graphic outline corresponding to the main graphic meet the preset fit condition. Ridge points are extracted from the pixels corresponding to the optimized sub-resolution auxiliary graphics, and the gradient value of the ridge point is higher than the gradient value of the other pixels in the neighborhood of the ridge point. The parameters are checked according to the mask rules, and the ridge lines are fitted using the ridge points; The ridge line is converted into a stepped broken line; A target sub-resolution auxiliary graphic is generated based on the stepped polyline, and the target sub-resolution auxiliary graphic is a Manhattan graphic. The step of generating the target sub-resolution auxiliary graphic based on the stepped polyline includes: Based on the stepped polyline, key inflection points are identified, and these key inflection points are connected to form a Manhattan graphic. The Manhattan graphic is then used as the target sub-resolution auxiliary graphic.

2. The method according to claim 1, characterized in that, The step of adjusting the sub-resolution auxiliary graphic corresponding to the main graphic to obtain an optimized sub-resolution auxiliary graphic includes: Set sub-resolution auxiliary graphic optimization evaluation points corresponding to the boundary of the main graphic, and the distance between the sub-resolution auxiliary graphic optimization evaluation points and the boundary of the main graphic satisfies a preset distance condition. The main graphic and the sub-resolution auxiliary graphic optimization evaluation points are input into a machine learning model for simulation calculation to obtain the exposure graphic contour of the main graphic. The simulation calculation is used to simulate the photolithography process and adjust the intensity value corresponding to the sub-resolution auxiliary graphic optimization evaluation points until the fit between the exposure graphic contour and the verification graphic contour corresponding to the main graphic meets the preset fit condition. The shape of the sub-resolution auxiliary graphic is adjusted based on the intensity value corresponding to the sub-resolution auxiliary graphic optimization evaluation point to obtain the optimized sub-resolution auxiliary graphic.

3. The method according to claim 1 or 2, characterized in that, The step of checking parameters according to mask rules and fitting ridge lines using the ridge points includes: According to the mask rule check parameters, ridge points with two adjacent spatial coordinate positions are grouped together to obtain multiple ridge point groups; Fit ridge points within the same ridge point group to a fitted line segment; By iterating through all the ridge point groups and connecting the fitted line segments in sequence, the ridge line is obtained.

4. The method according to claim 3, characterized in that, The step of converting the ridge line into a stepped broken line includes: The interval value is determined based on the mask rule check parameters, and the interval value is used to represent the interval distance between any two adjacent ridge points; Key points are extracted from the ridge points based on the interval values; Connect any adjacent key points with a polyline to transform the ridge line from a fitted line into a stepped polyline.

5. The method according to claim 4, characterized in that, The step of connecting any adjacent key points with a polyline to convert the ridge line from a fitted line into a stepped polyline includes: Starting from any key point, extend a first straight line until the first straight line intersects with a second straight line extending from a key point adjacent to that key point, thus obtaining a turning point. Connect the key points and the turning point in sequence to obtain the stepped broken line, wherein the first straight line and the second straight line are perpendicular to each other.

6. The method according to claim 5, characterized in that, The key inflection points include the turning points. The process of determining key inflection points based on the stepped polyline, connecting these key inflection points to form a Manhattan graphic, and defining the Manhattan graphic as the target sub-resolution auxiliary graphic includes: Extract the turning points in the stepped polyline, where the turning points are the endpoints of the polyline segments in the stepped polyline; Record the coordinates of the turning points to obtain the step points; Obtain the rectangle parameter from the mask rule check parameters, the rectangle parameter being used to determine the side length of the rectangle; The spatial coordinates of the rectangular point corresponding to the two adjacent step points are calculated based on the spatial coordinates of the two adjacent step points and the rectangular parameters to obtain the rectangular point. Starting from any rectangular point, connect the rectangular points sequentially with straight lines to generate the rectangles corresponding to the two adjacent step points; Iterate through all step points and generate rectangles corresponding to all step points; Starting from any rectangle, traverse all rectangles and connect the outer contours of the entire area covered by all rectangles in sequence to generate the Manhattan graphic, and determine the Manhattan graphic as the target sub-resolution auxiliary graphic.

7. A graphic extraction and optimization device, characterized in that, include: The acquisition module is used to acquire the main graphic from the mask design drawing file; The adjustment module is used to adjust the sub-resolution auxiliary graphic corresponding to the main graphic to obtain an optimized sub-resolution auxiliary graphic. The optimized sub-resolution auxiliary graphic is used to make the fit between the exposure graphic outline and the verification graphic outline corresponding to the main graphic meet the preset fit condition. The extraction module is used to extract ridge points from the pixels corresponding to the optimized sub-resolution auxiliary graphics, wherein the gradient value of the ridge point is higher than the gradient value of the other pixels in the neighborhood of the ridge point. The fitting module is used to check parameters according to mask rules and fit ridge lines using the ridge points; A conversion module is used to convert the ridge line into a stepped broken line; The generation module is used to generate a target sub-resolution auxiliary graphic based on the stepped polyline, wherein the target sub-resolution auxiliary graphic is a Manhattan graphic; The generation module is further configured to, based on the stepped polyline, determine key inflection points, connect the key inflection points to form a Manhattan graphic, and determine the Manhattan graphic as the target sub-resolution auxiliary graphic.

8. A graphic extraction and optimization device, characterized in that, The device includes: a processor and a memory storing computer program instructions; the processor, when executing the computer program instructions, implements the method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, implement the method as described in any one of claims 1-6.

10. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the electronic device, the electronic device causes the electronic device to perform the method as described in any one of claims 1-6.

Images