A method for repairing a lithography hotspot of a layout and a related apparatus

Abstract

This application discloses a method and related apparatus for repairing lithographic hotspots in a pattern. When executing the method provided in this application, a vector defect pattern with lithographic hotspots is first obtained from a mask pattern. This is then converted into a pixel defect pattern conforming to a target format, and a first set of slices is generated through sliding window slicing. Next, a pre-constructed hotspot geometric template is used to identify defective slices in the first set of slices. Based on hotspot image repair rules and the position of the defective slices, the pixel defect pattern is repaired and converted, ultimately yielding a vector-repaired pattern. This application, by converting the vector defect pattern into a pixel image and performing sliding window slicing, can automatically identify and repair lithographic hotspots, significantly reducing manual intervention and workload. Simultaneously, based on the pre-constructed hotspot geometric template and hotspot image repair rules, defective slices can be quickly and accurately identified and repaired, improving repair efficiency and effectiveness.

Description

Technical Field

[0001] This application relates to the field of integrated circuit manufacturing technology, specifically to a method and apparatus for repairing photolithographic hotspots in a layout. Background Technology

[0002] In integrated circuit mask patterns, the size of the critical layer pattern is often smaller than the wavelength of the lithography machine's light source. This prevents the lithography machine from achieving ideal imaging of the mask pattern, resulting in the Optical Proximity Effect (OPE). OPE causes distortion of the exposed pattern on the silicon wafer, such as rounded corners, linewidth deviations, shortened line ends, and missing areas, leading to defects like bridging or short lines. These areas that can cause wafer defects are called lithography hotspots. Once lithography hotspots appear, they can cause short circuits, open circuits, and poor contacts, severely affecting the electrical characteristics and manufacturing yield of integrated circuits. Traditional repair methods rely on manual experience and usually require manual repair, which is time-consuming and has poor convergence.

[0003] Therefore, how to effectively and quickly repair lithography hotspots caused by OPE, while avoiding the inefficiency and instability of traditional manual repair methods, is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a method and apparatus for repairing lithographic hotspots in a layout, which can effectively and quickly repair lithographic hotspot problems caused by OPE, while avoiding the inefficiency and instability of traditional manual repair methods.

[0005] The embodiments of this application disclose the following technical solutions:

[0006] A method for repairing lithographic hotspots in a layout, the method comprising:

[0007] Obtain the vector defect layout of the integrated circuit; the vector defect layout includes N mask patterns with photolithographic hot spots and M mask patterns without photolithographic hot spots, where N is a positive integer and M is a non-negative integer;

[0008] The vector defect layout is converted into a pixel image to obtain a pixel defect layout that conforms to the target format;

[0009] The pixel defect layout is sliced ​​using a sliding window process to obtain a first slice set; the first slice set includes multiple layout slices.

[0010] Defective slices in the first slice set are identified based on pre-constructed hotspot geometric templates;

[0011] Based on the hotspot image restoration rules and the position of the defect slice in the pixel defect map, the pixel defect map is restored and transformed to obtain a vector-restored map.

[0012] In one possible implementation, converting the vector defect layout into a pixel image to obtain a pixel defect layout conforming to the target format includes:

[0013] Identify the shape and vertex coordinates of each polygon in the vector defect layout;

[0014] Based on the shape and vertex coordinates of each polygon in the vector defect layout, the vector defect layout is edge-divided using Hanan meshing to obtain a meshed layout; the meshed layout includes multiple meshes.

[0015] The gridded layout is mapped to pixel space to obtain a Boolean matrix, and the horizontal and vertical spacing information between adjacent grids in the gridded layout is recorded.

[0016] By combining the Boolean matrix and the horizontal and vertical spacing information between adjacent grid images in the gridded layout, the vector defect layout is converted into the initial pixel layout;

[0017] The format of the initial pixel layout is converted to the target format to obtain the pixel defect layout.

[0018] In one possible implementation, the step of performing sliding window slicing on the pixel defect map to obtain a first slice set includes:

[0019] Define a slice window;

[0020] Using the slice window, starting from the upper left corner of the pixel defect map, slide along a preset step size from left to right and from top to bottom on the pixel defect map until a complete scan of the pixel defect map is completed, to obtain the first slice set;

[0021] In this process, the area covered by the slicing window each time during scanning is one of the layout slices, and all the layout slices constitute the first slice set.

[0022] In one possible implementation, the step of performing hotspot repair and transformation on the pixel defect map based on hotspot image inpainting rules and the position of the defect slice in the pixel defect map to obtain a vector-repaired map includes:

[0023] Based on the hotspot image restoration rules, hotspot restoration is performed on each of the defective slices to obtain multiple restored slices;

[0024] In the pixel defect layout, the image at each location of the defect slice is replaced with its repair slice to obtain the pixel repair layout;

[0025] The pixel-repaired pattern is vectorized to obtain the vector-repaired pattern.

[0026] In one possible implementation, the process of constructing the hotspot geometry template includes:

[0027] Select X sample layouts; the sample layouts include pixel-level mask patterns with multiple lithographic hotspots concentrated in the central region; X is a positive integer;

[0028] For each of the sample layouts, a sliding window slicing process is performed to obtain a second slice set corresponding to each sample layout; the second slice set includes multiple sample slices;

[0029] For each of the sample layouts, Y sample slices located around the central region of the sample layout are selected to obtain Y target slices; Y is a positive integer;

[0030] The X*Y target slices are merged and summarized to obtain the hotspot geometric template.

[0031] In one possible implementation, the hotspot image restoration rules include: moving graphic rules and shrinking graphic edge rules.

[0032] In one possible implementation, before vectorizing the pixel restoration pattern to obtain the vector restoration pattern, the method further includes:

[0033] The pixel repair layout is evaluated using a process window value. If the evaluation passes, the step of vectorizing the pixel repair layout to obtain the vector repair layout is performed.

[0034] A lithographic hotspot repair device for a layout, the device comprising:

[0035] An acquisition unit is used to acquire a vector defect layout of an integrated circuit; the vector defect layout includes a mask pattern with photolithographic hot spots;

[0036] The first image conversion unit is used to convert the vector defect layout into a pixel image to obtain a pixel defect layout that conforms to the target format;

[0037] The first slicing unit is used to perform sliding window slicing on the pixel defect layout to obtain a first slice set; the first slice set includes multiple layout slices.

[0038] A defect identification unit is used to identify defective slices in the first slice set based on a pre-built hotspot geometric template;

[0039] The first repair unit is used to perform hotspot repair and transformation on the pixel defect map based on hotspot image repair rules and the position of the defect slice in the pixel defect map to obtain a vector repair map.

[0040] A lithography hotspot repair device for a layout includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the lithography hotspot repair method for the layout as described above.

[0041] A computer-readable storage medium storing instructions that, when executed on a terminal device, cause the terminal device to perform the lithographic hotspot repair method for a layout as described above.

[0042] Compared with the prior art, this application has the following beneficial effects:

[0043] This application provides a method and related apparatus for repairing lithographic hotspots in a circuit layout. Specifically, when implementing the method for repairing lithographic hotspots in a circuit layout provided in this application, firstly, a vector defect layout of the integrated circuit is obtained: This step, by obtaining a vector defect layout containing lithographic hotspots, can comprehensively capture potential defects in the circuit design. This step can accurately identify the problem areas in the circuit design and provide necessary basic data for subsequent repair. Then, the vector defect layout is converted into a pixel image: By converting the vector layout into a pixel image, a pixel defect layout conforming to the target format is obtained. This conversion process can better adapt to the requirements of different lithography processes, laying the foundation for subsequent image processing and repair, and can effectively process different types of layout data. Next, the pixel defect layout is subjected to sliding window slicing: This step divides the pixel defect layout into multiple layout slices through sliding window slicing, thereby generating a first slice set. Each slice represents a local area of ​​the layout, facilitating detailed defect identification and repair of each slice, reducing processing complexity, and refining the analysis of defect areas. Building upon this foundation, defect slices are identified using a pre-constructed hotspot geometric template: this step utilizes the pre-constructed hotspot geometric template to quickly identify defect slices. This template recognition technology effectively improves the speed and accuracy of defect identification, avoiding errors and delays caused by manual inspection. Finally, repair is performed based on hotspot image repair rules: according to the hotspot image repair rules and the position of the defect slices in the pixel defect layout, hotspot repair and vector conversion are performed on the pixel defect layout to obtain a vector-repaired layout. This step effectively repairs hotspot problems generated during photolithography, ensuring that the repaired layout meets design standards and production requirements, significantly improving the quality and stability of the repair.

[0044] This application converts vector defect patterns into pixel images and performs sliding window slicing. It then automatically identifies defect slices using hotspot geometric templates, avoiding the inefficiency and instability of traditional manual repair processes. Furthermore, repair based on pre-constructed hotspot geometric templates and hotspot image repair rules allows for more precise repair of lithographic hotspots compared to manual methods, reducing errors. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in this embodiment or the prior art, the drawings used in the description of the embodiment or the prior art 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 A flowchart illustrating a method for repairing lithographic hotspots in a layout, as provided in this application embodiment;

[0047] Figure 2 A PNG format pixel defect layout provided for embodiments of this application;

[0048] Figure 3 A layout slice provided for an embodiment of this application;

[0049] Figure 4 A hotspot geometry template provided for embodiments of this application;

[0050] Figure 5 A flowchart illustrating a pixel defect layout conversion method provided in this application embodiment;

[0051] Figure 6 A schematic diagram of a Hanan mesh pixelation process provided in an embodiment of this application;

[0052] Figure 7 A Boolean matrix provided for embodiments of this application;

[0053] Figure 8 A flowchart illustrating another method for repairing lithographic hotspots in a layout provided in this application embodiment;

[0054] Figure 9 A schematic diagram of a correction target provided in an embodiment of this application;

[0055] Figure 10 A flowchart illustrating a method for constructing a hotspot geometry template, as provided in an embodiment of this application;

[0056] Figure 11 This is a schematic diagram of a lithographic hotspot repair device for a layout provided in an embodiment of this application. Detailed Implementation

[0057] To facilitate understanding of the technical solutions provided in the embodiments of this application, the background technology involved in the embodiments of this application will be described below.

[0058] "Poor convergence effect" means that traditional repair methods are not ideal or fail to quickly achieve the expected repair goals when solving lithography hotspot problems.

[0059] In integrated circuit manufacturing, the dimensions of each critical pattern in the designed mask pattern are smaller than the wavelength of the light source used in the lithography machine. The lithography machine forms the circuit pattern by illuminating the mask with a light source and projecting the pattern onto a silicon wafer through an optical system. However, the wavelength of the light source in a lithography machine is typically in the ultraviolet range, approximately tens to hundreds of nanometers (e.g., deep ultraviolet light at 193 nanometers). When the dimensions of critical patterns in the mask pattern (such as fine lines, holes, or other small structures) are smaller than the wavelength of the light source, imaging difficulties arise. Because the wavelength of light limits its resolution, it becomes impossible to accurately and completely transfer these tiny patterns onto the silicon wafer, resulting in optical proximity effect (OPE) and pattern distortion.

[0060] Currently, repairing lithographic hotspots typically relies on manual modifications by manufacturing engineers. This manual repair is not only time-consuming but also suffers from poor convergence. Meanwhile, with the continuous advancement of semiconductor technology nodes and the explosive growth of design rules, the number of lithographic hotspots generated in mask patterns is becoming increasingly massive, further slowing down manual repair.

[0061] To address this issue, this application provides a method and apparatus for repairing lithographic hotspots in a layout. First, a vector defect layout with lithographic hotspots is acquired and converted into a pixel defect layout conforming to a target format. Next, a sliding window slicing process is performed to obtain multiple layout slices. Using a pre-constructed hotspot geometric template, the defect slices are identified, and combined with hotspot repair rules and the positional information of the defect slices, the pixel defect layout is repaired and converted to obtain a repaired vector layout. This application avoids the inefficiency and instability of traditional manual repair methods by converting the vector defect layout into a pixel image and performing sliding window slicing, combined with the hotspot geometric template to automatically identify defect areas. Furthermore, automatic repair using a pre-designed hotspot geometric template and repair rules can repair lithographic hotspots more accurately than manual methods, reducing repair errors.

[0062] The technical solutions of the embodiments of the present invention 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 of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0063] See Figure 1 This figure is a flowchart of a method for repairing lithographic hotspots in a layout according to an embodiment of this application. Figure 1 As shown, the photolithographic hotspot repair method for this layout may include steps S101-S105:

[0064] S101: Obtain the vector defect layout of the integrated circuit.

[0065] In this photolithography hotspot repair method, a "vector defect layout" of the integrated circuit can first be obtained. This layout contains N mask patterns with photolithography hotspots and M mask patterns without photolithography hotspots. In other words, the vector defect layout... Figure 1 It includes mask patterns with photolithographic hotspots, but does not necessarily include mask patterns without photolithographic hotspots.

[0066] Vector defect layout: This is an integrated circuit layout described by vector data. It can accurately represent information such as the shape, size, and location of each mask pattern in the integrated circuit. During the photolithography process, some defects may occur, called "lithography hotspots," which may affect the performance or yield of the circuit.

[0067] Mask patterns with photolithographic hotspots: These patterns may develop defects during the photolithography process, usually due to improper layout design or problems with the photolithography process. These defects may cause the circuit to malfunction or even damage its functionality.

[0068] Mask patterns without photolithography hotspots: These patterns are normal parts that do not have defects during the photolithography process and do not require special repair treatment.

[0069] This step is fundamental to the entire repair method. Only by accurately obtaining the design layout with lithographic hotspots can subsequent identification and repair operations be carried out. By acquiring these vector layouts with defects, it is ensured that subsequent processing steps can accurately identify and repair the actual problems, thereby improving the design quality and manufacturing success rate of integrated circuits.

[0070] In one possible implementation, the image format for the vector defect layout is GDSII. GDSII (Graphic Data System II) is a standard format for describing integrated circuit layout data. A GDSII file is a binary file format used to store and transmit geometric information, hierarchical structure, and related attribute information of integrated circuit layouts.

[0071] S102: Convert the vector defect layout into a pixel image to obtain a pixel defect layout that conforms to the target format.

[0072] After obtaining the vector defect layout, the vector defect layout of the integrated circuit can be converted into a pixel image, thereby obtaining a pixel defect layout that conforms to the target format. Here, "vector defect layout" refers to a mask pattern containing lithographic hotspots, while "pixel image" is an image form composed of pixels that converts these vector data.

[0073] The purpose of this step is to facilitate subsequent processing and analysis. While vector data is precise, pixel images are more advantageous for large-scale image processing and machine learning model applications. By converting vector defect maps into pixel images, operations such as sliding window slicing and defect identification can be performed more easily. Furthermore, many existing image processing techniques and algorithms are better suited to processing pixelated image data; therefore, this step provides the necessary data format preparation for subsequent defect identification and repair processes.

[0074] In one possible implementation, the target format can be, but is not limited to, Portable Network Graphics (PNG) format. For example... Figure 2 As shown, Figure 2 A PNG format pixel defect layout is provided for an embodiment of this application.

[0075] S103: Perform sliding window slicing on the pixel defect layout to obtain the first slice set.

[0076] This step involves performing a sliding window slicing process on the previously obtained pixel defect layout to obtain the first slice set. Specifically, the sliding window slicing process involves using a fixed-size window to slide across the pixel defect layout at certain step sizes, cutting out multiple smaller layout slices from the layout. These slices constitute the first slice set.

[0077] This step decomposes the entire pixel defect pattern into multiple smaller, more easily processed slices for subsequent, more detailed analysis and processing. This method effectively improves the accuracy and efficiency of defect identification. Each slice can be independently matched with hotspot geometry templates and identified as a defect, which not only reduces computational resource consumption but also improves detection accuracy and avoids omissions or false positives. Figure 3 As shown, Figure 3 This is a layout slice provided for an embodiment of this application.

[0078] S104: Identify defective slices in the first slice set based on pre-built hotspot geometric templates.

[0079] After obtaining the set of slices from the pixel defect pattern (i.e., the first slice set), defective slices within this set can be identified based on a pre-constructed hotspot geometry template and defined as defective slices. This process involves matching a pre-defined geometry template (typically some known hotspot morphology or features) with multiple image slices extracted from the pixel defect pattern. Through this matching, slices that match the hotspot geometry template can be identified, i.e., defective slices with lithographic hotspots are discovered. Figure 4 As shown, Figure 4 A hotspot geometry template is provided for an embodiment of this application. The template consists of five sub-templates: (a), (b), (c), (d), and (e).

[0080] Specifically, a template matching method is used to compare the geometric template with the layout slice. When the similarity between the two is high, the layout slice is considered to contain a hotspot region. This application selects the Normalized Cross-Correlation (NCC) matching metric to evaluate the similarity between the layout slice and the geometric template. The formula for calculating the normalized cross-correlation is as follows:

[0081]

[0082] In the formula, NCC represents the similarity, and I(i) and T(i) are the i-th pixel values ​​of the layout slice and the geometric template, respectively. and Here, N represents the average pixel value of the layout slice and the geometric template, respectively, and N is the total number of pixels in the template. This formula calculates the normalized correlation coefficient (NCC) between the geometric template and the layout slice. The larger the NCC value, the higher the similarity. Next, the matching metric between the layout slice and the geometric template is calculated using this formula. A threshold is set; when the matching metric is higher than this threshold, the layout slice is considered to have extremely high similarity to the geometric template, meaning the layout slice contains hotspot areas. Finally, this layout slice is selected as the input data for the next repair stage (that is, to identify the defective slice in the first slice set).

[0083] This step aims to accurately locate and identify lithographic hotspots in the published image. By using hotspot geometric templates for matching, potential problem areas can be identified efficiently and accurately, providing clear targets for subsequent repair work. This method not only improves the accuracy of defect identification but also shortens processing time, as template matching algorithms can typically quickly identify slices that conform to specific patterns.

[0084] S105: Based on the hotspot image repair rules and the position of the defect slice in the pixel defect map, perform hotspot repair and transformation on the pixel defect map to obtain a vector repair map.

[0085] Finally, based on predefined hotspot image inpainting rules and the location of the identified defect slices within the pixel defect map, hotspot inpainting is performed on the original pixel defect map. Then, the inpainted pixel defect map is vector-converted to generate a vector-inpainted map. Specifically, this step involves the following points:

[0086] Hotspot Image Restoration Rules: These rules define how to process and correct detected lithographic hotspots. These rules may include specific algorithms or methods for eliminating or mitigating hotspot problems present in the layout.

[0087] Defect slice location: In step S104, defect slices with lithographic hotspots and their locations within the pixel defect pattern have been identified. This location information is crucial for accurately locating the areas requiring repair.

[0088] Pixel defect layout repair: Based on the above repair rules and location information, the original pixel defect layout is modified to eliminate lithographic hotspots. Then, the repaired pixel defect layout is vector-converted to generate a vector-repaired layout. The final generated vector-repaired layout will be an optimized layout, eliminating the original hotspot problem.

[0089] The purpose of this step is to practically correct lithographic hotspots in the layout, ensuring that the final layout avoids defects or errors caused by these hotspots during manufacturing. This correction process not only improves the quality of the layout but also enhances the success rate and reliability of integrated circuit manufacturing.

[0090] Based on the content of S101-S105, the process first involves acquiring a vector defect layout of an integrated circuit containing lithographic hotspots and converting it into a pixel defect layout conforming to the target format. Then, a sliding window slicing method is used to divide the pixel defect layout into multiple layout slices. Using a predefined hotspot geometric template, the layout slices containing defects are identified. Finally, based on hotspot repair rules and the location information of the defect slices, the pixel defect layout is repaired and converted, thereby generating a repaired vector layout. This application automates the process of converting vector defect layouts into pixel images, accurately identifying and repairing lithographic hotspots, thus improving repair efficiency and accuracy. Furthermore, the use of pre-constructed hotspot geometric templates and rules not only enables rapid detection of defect slices but also allows for precise location and repair of lithographic hotspots.

[0091] See Figure 5 , Figure 5 The flowchart of a pixel defect layout conversion method provided in this application embodiment shows that, correspondingly, step S102 converts the vector defect layout into a pixel image to obtain a pixel defect layout conforming to the target format, which can be specifically implemented through steps S501-S505:

[0092] S501: Identify the shape and vertex coordinates of each polygon in the vector defect layout.

[0093] To convert a vector defect layout into a pixel defect layout, the first step is to extract the specific shape of each polygon and its vertex coordinates from the vector defect layout. These polygons typically represent various regions in the integrated circuit layout, potentially including circuit elements. Identifying the shape and vertex coordinates of each polygon prepares the material for the subsequent meshing process.

[0094] In one possible implementation, identifying the shape and vertex coordinates of each polygon in the vector defect map includes:

[0095] Use Python's Klayout library to extract geometric data from a vector defect layout. This geometric data includes the shape of each polygon and the coordinates of its vertices in the vector defect layout.

[0096] It should be noted that the shapes of the polygons in the vector defect layout are generally regular rectangles.

[0097] S502: Based on the shape and vertex coordinates of each polygon in the vector defect map, perform edge-mapping Hanan meshing on the vector defect map to obtain a meshed map.

[0098] Hanan meshing is a method for converting vector layouts into mesh layouts. In this step, based on the shape of the polygons and their vertex coordinates, the vector layout is transformed into a meshed layout. Specifically, meshing maps the geometry of the vector defect layout onto a regular grid, and edge-setting processing makes the meshing result more accurate and standardized. The generated meshed layout contains multiple mesh maps, which help to accurately represent the defect layout.

[0099] This application introduces the Hanan grid instead of a uniformly distributed grid. The Hanan grid has a relatively flexible structure and can be dynamically adjusted according to the characteristics of the graphic and layout constraints. In areas with complex graphics, containing multiple corners and dense lines, the pixel size is set smaller; in areas with simple graphics and structures, the pixel size is set larger. The specific process of introducing the Hanan grid is described below: Figure 6 As shown, the grayscale graphic represents the shape within a slice of the map, composed of multiple polygons forming multiple edges. By extending these edges, the black dashed grid shown in the image can be obtained. These extended lines divide the slice graphic into a series of rectangular regions (i.e., a grid), each region representing one pixel.

[0100] S503: Map the gridded layout to pixel space to obtain a Boolean matrix, and record the horizontal and vertical spacing information between adjacent grids in the gridded layout.

[0101] After obtaining the gridded layout, it can be mapped to pixel space and converted into a Boolean matrix. Each element of the Boolean matrix corresponds to a pixel, and the value of the element indicates whether the pixel belongs to a defect area. This method converts the original vector-based data into a pixel-based representation. Furthermore, the horizontal and vertical spacing information is recorded, which helps to preserve the spatial relationships of the original layout when reconstructing it in pixel space.

[0102] Specifically, in Figure 6 Within each rectangular region, if a graphic exists, the pixel is encoded as 1; otherwise, it is encoded as 0. This yields... Figure 7 The Boolean matrix shown is used. Finally, the horizontal and vertical spacing between adjacent grids must be recorded. Combining the spacing information between Hanan grids with the Boolean matrix allows for a detailed and accurate representation of all graphic information and their relative positions within the layout slice.

[0103] S504: Combining the Boolean matrix and the horizontal and vertical spacing information between adjacent grid images in the gridded layout, the vector defect layout is converted into the initial pixel layout.

[0104] After obtaining the Boolean matrix, it can be combined with the spacing information in the meshed layout to convert the vector defect layout into an initial pixel layout. The Boolean matrix provides basic pixel information about the defective regions, while the recording of horizontal and vertical spacing helps ensure that spatial relationships are preserved when reconstructing the layout in pixel space.

[0105] In one possible implementation, converting the vector defect layout into the initial pixel layout by combining the Boolean matrix and the horizontal and vertical spacing information between adjacent grid images in the gridded layout includes:

[0106] Create an image canvas determined by the layout size and required resolution, and use the PIL library to combine the Boolean matrix with the horizontal and vertical spacing information between adjacent grid images in the gridded layout to draw geometry on the Hanan grid, accurately mapping the vector defect layout to pixel space.

[0107] S505: Convert the format of the initial pixel layout to the target format to obtain the pixel defect layout.

[0108] The final step is to convert the obtained initial pixel layout to the target format. This is to ensure that the generated pixel defect layout meets specific output requirements, which may include file format, resolution, or other specific format standards, ultimately generating a pixel defect layout that conforms to the target format.

[0109] Specifically, the completed map (i.e., the initial pixel map) is saved as an image file in the target format (such as PNG) to ensure the integrity of the graphic details and provide a foundation for subsequent hotspot repair operations. Figure 2 As shown, this process effectively converts vector defective layouts from GDSII format to PNG format by converting the vector layout into a pixelated image with a resolution of 120x120.

[0110] In one possible implementation, the step of performing sliding window slicing on the pixel defect map to obtain a first slice set includes:

[0111] A1 Define the slice window: First, you need to define the size and shape of a slice window. This window will cover a portion of the pixel defect pattern during subsequent scanning.

[0112] The scanning process of the A2 slice window: Starting from the top left corner of the pixel defect map, the slice window slides across the map along a preset step size. The sequence is: sliding from left to right until the right edge of the image is reached. Then it slides from top to bottom until the entire pixel defect map has been scanned.

[0113] In this context, the area covered by each slice window is considered a layout slice. Each slice is a small region that represents a part of the pixel defect layout.

[0114] It should be noted that all the map slices are acquired sequentially through a scanning process and eventually form a "first slice set". These slice sets contain multiple small blocks of the entire pixel defect map, and each slice contains information about different regions.

[0115] In one possible implementation, the size of the pixel defect pattern in this application can be 1.2 × 1.2 μm. 2 (square micrometers), therefore the size of the slice window can be set to 400×400nm. 2 (square nanometers), set the preset step size to 200nm (nanometers).

[0116] Specifically, the size of the pixel defect pattern in this application can be 1.2 × 1.2 μm. 2 To minimize the workload of subsequent processing while ensuring that the slices contain key features of the pattern, this application may choose to set the slice window size to 400×400nm. 2 Regarding the selection of the preset step size, if the sliding step size is the same as the window size, the graphics located at the boundaries of two adjacent windows will be segmented into different slice graphics, resulting in the loss of some key graphic structures and affecting subsequent repair operations. To avoid this situation, the step size is generally smaller than the slice window size. However, a smaller step size will result in too many slices being generated, producing a large amount of redundant information and placing greater processing and computational pressure on subsequent steps. Therefore, this application sets the step size to 200nm. Converting the above settings to the image format, the layout image resolution is 120×120, the slice window size is 40×40 pixels, and the preset step size is 20 pixels. Based on this parameter setting, each layout sample can obtain 25 layout slices. Following the above sliding window method... Figure 2 The displayed map sample has been sliced, and one of the map slices is shown below. Figure 3 As shown.

[0117] This process divides a large image into multiple smaller slices using a slicing window, in order to facilitate subsequent defect identification or repair.

[0118] See Figure 8 , Figure 8The flowchart of another method for photolithographic hotspot repair of a layout provided in this application embodiment shows that, correspondingly, step S105, based on the hotspot image repair rules and the position of the defect slice in the pixel defect layout, performs hotspot repair and conversion on the pixel defect layout to obtain a vector repair layout, which can be specifically implemented through steps S801-S803:

[0119] S801: Based on the hotspot image repair rules, perform hotspot repair on each of the defective slices to obtain multiple repair slices.

[0120] In step S801, these repair rules are first applied to each defective slice, and each repaired defective slice is called a repaired slice.

[0121] In one possible implementation, image preprocessing, contour detection, and feature size calculation are required for each of the defective slices before hotspot repair is performed.

[0122] Specifically, to repair defective slices containing hotspot regions, it is first necessary to extract information such as the boundary coordinates and feature dimensions of the defective slices, which requires the application of knowledge from the field of image processing. Specifically, the process of extracting this information consists of two main stages: image preprocessing and contour detection. First, image preprocessing is the foundation of the entire repair operation. It includes converting the original color image to a grayscale image and binarization. Color images contain multiple channels, while grayscale images contain only a single brightness channel. This conversion can be achieved using the `cvtColor()` function in the Open Source Computer Vision Library (OpenCV). Grayscale conversion simplifies the image complexity, making subsequent processing more efficient. Next, binarization uses the `threshold()` function to convert the grayscale image into a binary image with only black and white. This step aims to highlight the difference between the image and the background, making the image contours clearer and laying the foundation for the next step of contour detection.

[0123] After image preprocessing, the next step is contour detection, the main goal of which is to identify all closed boundaries in the image and extract their coordinates. In OpenCV, the `findContours()` function is the core tool for achieving this. It scans the binary image, detecting and extracting all contours based on pixel connectivity. Each contour consists of a series of consecutive coordinate points representing the boundary information of the shape. The `findContours()` function returns a list of contours, where each contour is an array containing pixel coordinates. Through this process, the boundary information of the shapes in the layout slices is extracted. This information is crucial for subsequent analysis and repair, determining the accuracy of the repair operation.

[0124] After completing contour detection and extracting the boundary coordinate information of the layout slices, the next step is to calculate the feature dimensions of the graphics in the defect slices based on this information. The calculation methods for feature dimensions differ depending on the type of graphic. First, the detected contours need to be classified, which can be achieved by analyzing their geometric and topological features. For example, by calculating the number of endpoints, side length, boundary rectangle, aspect ratio, and minimum circumcircle of each contour, we can identify line graphics with high aspect ratios, independent cubes with aspect ratios close to 1, and corner graphics with obvious corner points. The identification of line endpoints and T-shaped structures can be determined by analyzing the number of endpoints and branching structures of the contour. Based on these classification results, the feature dimensions of each graphic can be further calculated as a basis for subsequent repair operations.

[0125] S802: In the pixel defect layout, the image at each location of the defect slice is replaced with its repair slice to obtain the pixel repair layout.

[0126] Step S802 involves applying the repaired defect slices to the corresponding positions in the original pixel defect map. Specifically, the repaired version of each defect slice replaces the original defective image area.

[0127] This process is performed at the pixel level. The repaired slices are replaced one by one in the pixel defect pattern to ensure that each repaired area is correctly supplemented and adjusted. The result is a new pixel repair pattern that repairs the original defect and perfectly matches the repaired slices at the pixel level.

[0128] S803: Vectorize the pixel restoration pattern to obtain the vector restoration pattern.

[0129] Step S803 involves converting the repaired pixel-based layout back into vector form. Vectorization is the process of converting pixel-based image data back into a geometry-based representation, which is particularly important for integrated circuit layouts because vector data can accurately describe circuit structures and interconnections.

[0130] In this process, the repaired pixel layout will be analyzed and converted into vector data, ultimately generating a vector repair layout. This vector repair layout contains the repaired defective areas and conforms to the format requirements of vector layouts, which can be used for subsequent production or design verification.

[0131] In one possible implementation, the hotspot image restoration rules include: moving graphic rules and shrinking graphic edge rules.

[0132] Specifically, the moving graphic rule increases the spacing between graphics by translating them as a whole. The steps involve identifying the graphic to be moved and then translating it a certain step size in a specified direction to ensure that the spacing is increased without changing the graphic's size. The shrinking graphic edge rule, on the other hand, precisely controls local spacing by adjusting a specific edge of the graphic. The steps involve identifying the edge that needs adjustment and then shrinking that edge in the desired direction to increase the spacing. This method is particularly suitable for local optimization, as it can repair hotspots and reduce the generation of new hotspots without affecting other areas.

[0133] In one possible implementation, this application defines a function `shrink_pattern(contour, side, step)` to shrink (or move) the edges of a graphic with a fixed step size. Here, `contour` represents the contour to be processed, which is an array containing two-dimensional pixel coordinates; `side` represents the edge to be shrunk; and `step` represents the shrinking step size, i.e., the distance the edge is moved along a specified direction. The function first obtains the coordinates of the boundary points of the contour, then uses the `np.clip()` function to adjust the coordinates of the corresponding boundaries, moving each edge by a fixed step size while ensuring that the adjusted coordinates do not exceed the range of the initial contour. Next, it selects the specified edge (top, bottom, left, right) for the movement operation based on the input `side` parameter. Finally, the modified contour is output as the function's return value.

[0134] The key to the repair operation lies in how to reasonably select the parameters of the above functions. This applies to the five geometric templates containing hotspot regions obtained earlier (such as...). Figure 4 As shown), the rules for repairing geometric templates (i.e., the selection of the shapes to be repaired and the edges to be moved) are specified in different geometric templates. Figure 9As shown, the thick black solid lines represent the targets that need correction. According to this rule, for hotspot areas belonging to different templates, the contours that need repair and the edges that need to be moved can be selected, and then used as the `contour` and `side` parameters in the `shrink_pattern()` function.

[0135] Next, we will discuss how to select an appropriate step. In the function described above, the setting of the step parameter has a significant impact on the process window and repair time. A larger step can speed up the repair process and reduce the repair time required because each adjustment is more significant, thus reducing the number of iterations needed. However, this can also lead to over-adjustment or insufficient precision, resulting in a poor process window for the repaired layout. Conversely, a smaller step, while increasing the repair time overhead, allows for more refined corrections each time, resulting in a better process window for the repaired layout. Therefore, a balance needs to be struck between repair time and ensuring good performance to achieve efficient layout repair. For this reason, in applications... Figure 4 When applying the repair rules from (a) to (d), the function's step is set to 2; while in the application... Figure 4 In rule (e), set step to 1. Note that 2 and 1 here refer to the image pixels, which, when converted to GDSII format, correspond to 0.02um and 0.01um respectively.

[0136] In one possible implementation, before vectorizing the pixel restoration pattern in step S803 to obtain the vector restoration pattern, the method further includes B1-B2:

[0137] B1: Process Window Value Evaluation: Before performing vectorization, the process window value of the pixel restoration layout needs to be evaluated. This step checks whether the restored pixel layout meets certain process requirements or standards. If the process window value evaluation is passed, meaning the restored layout meets the predetermined requirements, then vectorization processing can proceed.

[0138] B2: Perform vectorization: Once the evaluation is passed, the step of vectorizing the pixel restoration layout can be performed to obtain the vector restoration layout.

[0139] See Figure 10 , Figure 10 A flowchart illustrating a method for constructing a hotspot geometry template provided in this application embodiment is shown, which can be specifically implemented through steps S1001-S1004:

[0140] S1001: Select X sample layouts.

[0141] In step S1001, X sample layouts need to be selected, where the pixel-level mask patterns of the lithographic hotspots in these layouts are concentrated in the central area of ​​the layout.

[0142] Sample layouts refer to hot spots (areas prone to failure or performance issues) that appear in the central region of the layout during the manufacturing process due to different parameters or defects in the photolithography process. Therefore, these sample layouts are used to simulate and analyze the distribution characteristics of hot spots.

[0143] X is a positive integer, meaning that multiple sample layouts can be selected to improve the reliability and adaptability of the data.

[0144] S1002: Perform sliding window slicing for each of the sample layouts to obtain a second slice set corresponding to each of the sample layouts.

[0145] In step S1002, a sliding window slicing process is performed on each selected sample layout. Sliding window slicing refers to applying a sliding window algorithm to divide the sample layout into multiple small regions, i.e., slices, according to a certain size and step size.

[0146] The second slice set is a collection of multiple sample slices obtained through sliding window slicing. Each sample slice is a small region of the map, containing hotspot information around the central region.

[0147] The purpose of this process is to extract smaller slices from each sample map to facilitate subsequent analysis and processing.

[0148] It should be noted that the sliding window slicing process in step S1002 is similar to the sliding window slicing process in step S103, and will not be described again here.

[0149] S1003: For each of the sample layouts, select Y sample slices located around the central region of the sample layout to obtain Y target slices.

[0150] In step S1003, for each sample layout, Y sample slices located around the central region of the layout are selected from the obtained second slice set.

[0151] Y is a positive integer representing the selection of multiple slices from around the central region. These slices are the most representative of the hotspot area. By selecting these slices, the geometric features of the hotspots can be captured in a concentrated manner.

[0152] These target slices will be used for further template construction and analysis.

[0153] S1004: Merge and summarize the X*Y target slices to obtain the hotspot geometric template.

[0154] In step S1004, all target slices obtained from step S1003 are merged and summarized.

[0155] The process of merging and summarizing involves extracting common features from different slices to obtain hotspot geometric templates (such as...). Figure 4 As shown, Figure 4 The hotspot geometry templates include five sub-templates: (a), (b), (c), (d), and (e). These templates represent common hotspot geometric features around the central region and can be used for subsequent repairs.

[0156] These hotspot geometry templates can help identify potential hotspot areas, allowing for remedial measures to be taken.

[0157] For example, this application uses cluster analysis to analyze sample maps containing hotspots to obtain geometric templates representing the design rules of hotspot areas. For sample maps containing hotspots, the hotspots are located in the central region of the map. Therefore, in order to reduce the computational burden of subsequent cluster analysis, Y sample slices (e.g., 12 sample slices) located around the central region of the map are selected from the sample slices (e.g., 25 sample slices) of each map sample, resulting in X*Y target slices (if X is 100, X*Y = 1200). Then, cluster analysis is used to gradually merge the nearest slices among the 1200 target slices until they cannot be merged, thereby dividing the target slices containing hotspots into 5 clusters.

[0158] After performing cluster analysis on the target slices, a set of geometric rules is summarized from each cluster to represent the geometric characteristics of hotspot regions in all slices within that group. These geometric rules consist of basic graphic structures in the layout (such as line endpoints, corners, dense lines, etc.) or combinations of some of these graphics. Each cluster yields a corresponding set of geometric rules, i.e., a geometric template. Ultimately, five geometric templates are obtained for use in this application, such as... Figure 4 As shown.

[0159] In one possible implementation, the Agglomerative Nesting (AGNES) algorithm can be used to perform cluster analysis on the target slice. AGNES is a typical clustering algorithm for this purpose. This algorithm calculates the distance between data points (usually Euclidean distance or other distance metrics) and then progressively merges or clusters data points based on the principle of closest proximity, forming a hierarchical cluster tree. The average distance of all samples in two sample sets is selected as the inter-cluster distance, while the merge distance refers to the standard distance used to select two clusters to merge during the clustering process; it is actually used to determine which two clusters to merge. Then, the inflection point method is used to select an appropriate merge distance.

[0160] See Figure 11 , Figure 11 This is a schematic diagram of a lithographic hotspot repair device for a layout provided in an embodiment of this application. Figure 11 As shown, the photolithographic hotspot repair device for this layout includes:

[0161] The acquisition unit 1101 is used to acquire the vector defect layout of the integrated circuit; the vector defect layout includes a mask pattern with photolithographic hot spots;

[0162] The first image conversion unit 1102 is used to convert the vector defect layout into a pixel image to obtain a pixel defect layout that conforms to the target format;

[0163] The first slicing unit 1103 is used to perform sliding window slicing processing on the pixel defect layout to obtain a first slice set; the first slice set includes multiple layout slices.

[0164] The defect identification unit 1104 is used to identify defective slices in the first slice set based on a pre-built hotspot geometric template;

[0165] The first repair unit 1105 is used to perform hotspot repair and transformation on the pixel defect map based on hotspot image repair rules and the position of the defect slice in the pixel defect map to obtain a vector repair map.

[0166] In one possible implementation, the first image conversion unit 1102 specifically includes:

[0167] The identification unit is used to identify the shape and vertex coordinates of each polygon in the vector defect layout;

[0168] A meshing unit is used to perform Hanan meshing on the vector defect layout based on the shape and vertex coordinates of each polygon in the vector defect layout to obtain a meshed layout; the meshed layout includes multiple meshes;

[0169] The mapping unit is used to map the gridded layout to the pixel space to obtain a Boolean matrix, and to record the horizontal and vertical spacing information between adjacent grids in the gridded layout;

[0170] The second image conversion unit is used to combine the Boolean matrix and the horizontal and vertical spacing information between adjacent grid images in the gridded layout to convert the vector defect layout into the initial pixel layout;

[0171] The format conversion unit is used to convert the format of the initial pixel layout into the target format to obtain the pixel defect layout.

[0172] In one possible implementation, the first slicing unit 1103 specifically includes:

[0173] Define a unit, used to define a slice window;

[0174] The scanning unit is used to use the slicing window to slide from the upper left corner of the pixel defect map along a preset step size from left to right and from top to bottom on the pixel defect map until the complete scan of the pixel defect map is completed, thereby obtaining the first slice set.

[0175] In this process, the area covered by the slicing window each time during scanning is one of the layout slices, and all the layout slices constitute the first slice set.

[0176] In one possible implementation, the first repair unit 1105 specifically includes:

[0177] The second repair unit is used to perform hotspot repair on each of the defective slices based on hotspot image repair rules to obtain multiple repair slices.

[0178] The replacement unit is used to replace the image at each location of the defective slice in the pixel defect layout with its repair slice to obtain a pixel repair layout.

[0179] The vectorization unit is used to vectorize the pixel restoration pattern to obtain the vector restoration pattern.

[0180] In one possible implementation, the device further includes:

[0181] The layout selection unit is used to select X sample layouts; the sample layouts include pixel-level mask patterns in which multiple lithographic hotspots are concentrated in the central region; X is a positive integer;

[0182] The second slicing unit is used to perform sliding window slicing processing on each of the sample layouts to obtain a second slice set corresponding to each of the sample layouts; the second slice set includes multiple sample slices;

[0183] The slice selection unit is used to select Y sample slices located around the central region of each sample layout to obtain Y target slices; Y is a positive integer.

[0184] The merging and summarizing unit is used to merge and summarize the X*Y target slices to obtain the hotspot geometric template.

[0185] In one possible implementation, the hotspot image restoration rules include: moving graphic rules and shrinking graphic edge rules.

[0186] In one possible implementation, the device further includes:

[0187] An evaluation unit is used to evaluate the process window value of the pixel repair layout;

[0188] If the evaluation passes, the execution unit is used to perform the step of vectorizing the pixel restoration pattern to obtain the vector restoration pattern.

[0189] In addition, this application embodiment also provides a lithographic hotspot repair device for a layout, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the lithographic hotspot repair method for the layout as described above.

[0190] In addition, this application embodiment also provides a computer-readable storage medium storing instructions that, when executed on a terminal device, cause the terminal device to perform the lithographic hotspot repair method for the layout as described above.

[0191] This application provides a lithographic hotspot repair apparatus for a layout. First, an acquisition unit 1101 acquires a vector defect layout of the integrated circuit. Then, a first image conversion unit 1102 converts the vector defect layout into a pixel image to obtain a pixel defect layout conforming to a target format. A first slicing unit 1103 performs sliding window slicing on the pixel defect layout to obtain a first slice set, which includes multiple layout slices. Next, a defect identification unit 1104 identifies the defect slices in the first slice set based on a pre-constructed hotspot geometric template. This allows a first repair unit 1105 to repair and convert the pixel defect layout based on hotspot image repair rules and the position of the defect slices in the pixel defect layout to obtain a vector repair layout. This application achieves automatic identification and repair of lithographic hotspots by converting the vector defect layout into a pixel image and performing sliding window slicing, significantly reducing manual intervention and workload. Simultaneously, relying on pre-constructed hotspot geometric templates and repair rules, defect areas can be quickly and accurately identified and repaired, thereby improving repair efficiency and effectiveness.

[0192] The foregoing has provided a detailed description of a method and related apparatus for repairing lithographic hotspots in a layout, as provided in this application. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

[0193] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0194] It should also be noted that, in this application, 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 one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A method for repairing photolithographic hotspots in a layout, characterized in that, The method includes: Obtain the vector defect layout of the integrated circuit; the vector defect layout includes N mask patterns with photolithographic hot spots and M mask patterns without photolithographic hot spots, where N is a positive integer and M is a non-negative integer; The vector defect layout is converted into a pixel image to obtain a pixel defect layout that conforms to the target format; The pixel defect layout is sliced ​​using a sliding window process to obtain a first slice set; the first slice set includes multiple layout slices. Defective slices in the first slice set are identified based on pre-constructed hotspot geometric templates; Based on the hotspot image restoration rules and the position of the defect slice in the pixel defect map, the pixel defect map is restored and transformed to obtain a vector-restored map.

2. The method according to claim 1, characterized in that, The step of converting the vector defect layout into a pixel image to obtain a pixel defect layout conforming to the target format includes: Identify the shape and vertex coordinates of each polygon in the vector defect layout; Based on the shape and vertex coordinates of each polygon in the vector defect layout, the vector defect layout is edge-divided using Hanan meshing to obtain a meshed layout; the meshed layout includes multiple meshes. The gridded layout is mapped to pixel space to obtain a Boolean matrix, and the horizontal and vertical spacing information between adjacent grids in the gridded layout is recorded. By combining the Boolean matrix and the horizontal and vertical spacing information between adjacent grid images in the gridded layout, the vector defect layout is converted into the initial pixel layout; The format of the initial pixel layout is converted to the target format to obtain the pixel defect layout.

3. The method according to claim 1, characterized in that, The first slice set is obtained by performing sliding window slicing on the pixel defect layout, including: Define a slice window; Using the slice window, starting from the upper left corner of the pixel defect map, slide along a preset step size from left to right and from top to bottom on the pixel defect map until a complete scan of the pixel defect map is completed, to obtain the first slice set; In this process, the area covered by the slicing window each time during scanning is one of the layout slices, and all the layout slices constitute the first slice set.

4. The method according to claim 1, characterized in that, The step of performing hotspot repair and transformation on the pixel defect map based on hotspot image repair rules and the position of the defect slice in the pixel defect map to obtain a vector repair map includes: Based on the hotspot image restoration rules, hotspot restoration is performed on each of the defective slices to obtain multiple restored slices; In the pixel defect layout, the image at each location of the defect slice is replaced with its repair slice to obtain the pixel repair layout; The pixel-repaired pattern is vectorized to obtain the vector-repaired pattern.

5. The method according to claim 1, characterized in that, The process of constructing the hotspot geometry template includes: Select X sample layouts; the sample layouts include pixel-level mask patterns with multiple lithographic hotspots concentrated in the central region; X is a positive integer; For each of the sample layouts, a sliding window slicing process is performed to obtain a second slice set corresponding to each sample layout; the second slice set includes multiple sample slices; For each of the sample layouts, Y sample slices located around the central region of the sample layout are selected to obtain Y target slices; Y is a positive integer; The X*Y target slices are merged and summarized to obtain the hotspot geometric template.

6. The method according to any one of claims 1-5, characterized in that, The hotspot image restoration rules include: moving graphic rules and shrinking graphic edge rules.

7. The method according to claim 4, characterized in that, Before vectorizing the pixel restoration pattern to obtain the vector restoration pattern, the method further includes: The pixel repair layout is evaluated using a process window value. If the evaluation passes, the step of vectorizing the pixel repair layout to obtain the vector repair layout is performed.

8. A device for repairing lithographic hotspots in a layout, characterized in that, The device includes: An acquisition unit is used to acquire a vector defect layout of an integrated circuit; the vector defect layout includes a mask pattern with photolithographic hot spots; The first image conversion unit is used to convert the vector defect layout into a pixel image to obtain a pixel defect layout that conforms to the target format; The first slicing unit is used to perform sliding window slicing on the pixel defect layout to obtain a first slice set; the first slice set includes multiple layout slices. A defect identification unit is used to identify defective slices in the first slice set based on a pre-built hotspot geometric template; The first repair unit is used to perform hotspot repair and transformation on the pixel defect map based on hotspot image repair rules and the position of the defect slice in the pixel defect map to obtain a vector repair map.

9. A lithographic hotspot repair device for a layout, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the lithographic hotspot repair method for a layout as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on a terminal device, cause the terminal device to perform the lithographic hotspot repair method for a layout as described in any one of claims 1-7.

Images