Multi-exposure mask image set generation method and device, equipment and storage medium

Abstract

The invention relates to the technical field of chip manufacturing, and discloses a multi-exposure mask image set generation method and device, equipment and a storage medium, and the method comprises the steps: obtaining an initial mask image; inputting the initial mask image into a pre-trained diffusion model to carry out noise adding processing and de-noising processing; selecting a plurality of step numbers of de-noising processing as selected step numbers; respectively calculating an accumulated image difference between every two adjacent selected step numbers, and respectively generating a first intermediate mask image corresponding to each accumulated image difference; superposing the noise image to any one of the first intermediate mask images to generate a superposed intermediate mask image, and updating the superposed original first intermediate mask image into the superposed intermediate mask image to serve as a second intermediate mask image; and carrying out pixel-by-pixel binarization processing on each second intermediate mask image, and taking the image subjected to binarization processing as a final mask image, so that automatic splitting of the initial mask image is realized.

Description

Technical Field

[0001] The present invention relates to the field of chip manufacturing technology, and in particular to a method, device, equipment and storage medium for generating a multiple exposure mask image set. Background Art

[0002] In the field of semiconductor chip manufacturing, photolithography is one of the core processes for achieving integrated circuit patterning. As chip feature sizes continue to shrink, especially as they enter the 3-nanometer (3nm) and below process nodes, traditional photolithography faces unprecedented challenges. The realization of this process level is highly dependent on extreme ultraviolet (EUV) lithography machines, which can provide higher resolution to meet the manufacturing needs of precision patterns.

[0003] In order to overcome the problem of equipment accuracy, the industry has begun to explore the use of multiple exposure technology as a potential alternative. Multiple exposure technology accumulates the required high-precision features through multiple patterning processes on an immersion deep ultraviolet (DUV) lithography machine. This method has been verified in 5-nanometer (5nm) chip manufacturing and has the potential to be extended to 3nm and below process nodes in theory.

[0004] In the implementation of multiple exposure technology, mask generation is a crucial link. The quality of the mask directly affects the pattern accuracy and yield of the final chip. However, the current research on multiple exposure mask generation is still insufficient. In most cases, the splitting of the mask relies on manual operation, which is not only inefficient, but also lacks unified standards, resulting in differences between masks produced in different batches or different factories, making it difficult to ensure the consistency and stability of the process. Summary of the invention

[0005] Based on this, it is necessary to propose a method, device, equipment and storage medium for generating a re-exposure mask image set in order to address the problem that the existing splitting of multiple exposure masks relies on manual labor, resulting in low efficiency and lack of stability.

[0006] A first aspect of the present invention provides a method for generating a multi-exposure mask image set, the method comprising:

[0007] Obtain an initial mask image;

[0008] The initial mask image is input into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0009] Selecting a number of steps of the denoising process as selected steps according to a preset method;

[0010] According to the image difference of each step of removing noise information, the accumulated image difference between every two adjacent selected steps is calculated respectively, and a first intermediate mask image corresponding to each accumulated image difference is generated respectively, and a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0011] The noise image is superimposed on any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and the superimposed original first intermediate mask image is updated to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0012] Each of the second intermediate mask images is subjected to a pixel-by-pixel binarization process, and the binarized image is used as a final mask image. A plurality of the final mask images constitute the multiple exposure mask image set.

[0013] Furthermore, the step of inputting the initial mask image into a pre-trained diffusion model for denoising further includes:

[0014] The diffusion model is used to gradually remove noise information from the noise image, and the information difference of the noise information removed at each step is obtained.

[0015] Furthermore, the step of selecting a number of steps of the denoising process as selected steps in a preset manner includes:

[0016] Set the intermediate information difference threshold;

[0017] From the steps of the denoising process, a number of selected steps are determined so that the cumulative information difference between adjacent selected steps is within the intermediate information difference threshold; wherein the cumulative information difference between adjacent selected steps is obtained by superimposing the information difference of each step between adjacent selected steps after removing the noise information.

[0018] Furthermore, the step of performing pixel-by-pixel binarization on each of the second intermediate mask images and using the binarized images as the final mask images includes:

[0019] Comparing the pixel values ​​of the first designated pixels at the same position in each of the second intermediate mask images to obtain a first maximum pixel value;

[0020] Setting the pixel value of the first designated pixel in the second intermediate mask image corresponding to the first maximum pixel value to 1, and setting the pixel values ​​of the first designated pixels in the remaining second intermediate mask images to 0;

[0021] Each image after traversing all pixels is used as the final mask image.

[0022] Furthermore, the step of using each image after traversing all pixels as the final mask image includes:

[0023] Acquire an image difference between the target image and the initial mask image;

[0024] Obtaining the position of the difference pixel point corresponding to the image difference;

[0025] The pixel values ​​of non-zero pixels of the difference pixels in the final mask image are set to 0 to obtain an updated final mask image, and the updated final mask image is used to replace the original image in the multiple exposure mask image set.

[0026] Furthermore, the step of performing pixel-by-pixel binarization on each of the second intermediate mask images and using the binarized images as the final mask images includes:

[0027] Acquire a pixel position corresponding to an exposed photolithography image in the initial mask image, and record each pixel corresponding to the exposed photolithography image as a second designated pixel;

[0028] setting the pixel value of the pixel at a different pixel position from the second designated pixel in each of the second intermediate mask images to 0;

[0029] Compare pixel values ​​at the same pixel position as the second designated pixel in each of the second intermediate mask images to obtain a second maximum pixel value;

[0030] Setting the pixel value of the pixel position corresponding to the second maximum pixel value in the second intermediate mask image corresponding to the second maximum pixel value to 1; and setting the pixel values ​​of the pixel positions corresponding to the second maximum pixel value in the remaining second intermediate mask images to 0;

[0031] Each image after traversing all pixels is used as the final mask image.

[0032] Furthermore, the step of selecting a number of steps of the denoising process as selected steps in a preset manner includes:

[0033] Set the starting number of steps and step interval;

[0034] The step interval is superimposed on the starting step number, and the number of steps obtained by superimposing the step interval each time is used as the selected step number, until the superimposed value exceeds the maximum number of steps for denoising processing.

[0035] A second aspect of the present invention provides a device for generating a multi-exposure mask image set, the device comprising:

[0036] An acquisition module, used for acquiring an initial mask image;

[0037] A noise processing module is used to input the initial mask image into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0038] A selection module, used for selecting a number of steps of the denoising process as selected steps according to a preset method;

[0039] A first image set generating module, configured to calculate the accumulated image difference between every two adjacent selected steps according to the image difference of each step of removing noise information, and to generate a first intermediate mask image corresponding to each accumulated image difference, wherein a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0040] A second image set generating module, configured to superimpose the noise image onto any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and update the superimposed original first intermediate mask image to the superimposed intermediate mask image, so that the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0041] The mask image set generation module is used to perform pixel-by-pixel binarization on each of the second intermediate mask images, and use the binarized image as the final mask image. Several of the final mask images constitute the multiple exposure mask image set.

[0042] A third aspect of the present invention provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor performs the following steps:

[0043] Obtain an initial mask image;

[0044] The initial mask image is input into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0045] Selecting a number of steps of the denoising process as selected steps according to a preset method;

[0046] According to the image difference of each step of removing noise information, the accumulated image difference between every two adjacent selected steps is calculated respectively, and a first intermediate mask image corresponding to each accumulated image difference is generated respectively, and a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0047] The noise image is superimposed on any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and the superimposed original first intermediate mask image is updated to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0048] Each of the second intermediate mask images is subjected to a pixel-by-pixel binarization process, and the binarized image is used as a final mask image. A plurality of the final mask images constitute the multiple exposure mask image set.

[0049] A fourth aspect of the present invention provides a computer device, the device comprising a memory and a processor, the memory storing a computer program, and when the computer program is executed by the processor, the processor performs the following steps:

[0050] Obtain an initial mask image;

[0051] The initial mask image is input into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0052] Selecting a number of steps of the denoising process as selected steps according to a preset method;

[0053] According to the image difference of each step of removing noise information, the accumulated image difference between every two adjacent selected steps is calculated respectively, and a first intermediate mask image corresponding to each accumulated image difference is generated respectively, and a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0054] The noise image is superimposed on any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and the superimposed original first intermediate mask image is updated to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0055] Each of the second intermediate mask images is subjected to a pixel-by-pixel binarization process, and the binarized image is used as a final mask image. A plurality of the final mask images constitute the multiple exposure mask image set.

[0056] The present invention performs denoising and denoising on the initial mask image by using a pre-trained diffusion model, extracts a series of effective first intermediate mask images from the denoising process, and binarizes the first intermediate mask images to obtain a final mask image set, thereby realizing automatic segmentation of the initial mask image and obtaining a set of multiple-exposure mask images without relying on manual experience for segmentation, greatly improving efficiency, and the results are reliable and stable. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] in:

[0059] Figure 1A diagram of an application environment of a method for generating a multi-exposure mask image set in one embodiment;

[0060] Figure 2 is a flow chart of a method for generating a multi-exposure mask image set in one embodiment;

[0061] Figure 3 is a structural block diagram of a device for generating a multiple exposure mask image set in one embodiment;

[0062] Figure 4 FIG. 4 is a structural block diagram of a computer device in one embodiment. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] Figure 1 An application environment map is generated for a set of multiple exposure mask images in one embodiment. Figure 1 , the multiple exposure mask image set generation method is applied to a multiple exposure mask image set generation system. The multiple exposure mask image set generation system includes a terminal 110 and a server 120. The terminal 110 and the server 120 are connected via a network, and the terminal 110 can be a desktop terminal or a mobile terminal, and the mobile terminal can be at least one of a mobile phone, a tablet computer, a laptop computer, etc. The server 120 can be implemented as an independent server or a server cluster composed of multiple servers. The terminal 110 is used to execute the steps of the multiple exposure mask image set generation method, and the server 120 is used to store the diffusion model used in the multiple exposure mask image set generation method.

[0065] like Figure 2 As shown, in one embodiment, a method for generating a multi-exposure mask image set is provided. The method can be applied to a terminal or a server. This embodiment is described by applying to a terminal as an example. The method for generating a multi-exposure mask image set specifically includes the following steps:

[0066] S1: Get the initial mask image;

[0067] S2: inputting the initial mask image into a pre-trained diffusion model for noise addition and denoising; the noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image; the denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference of removing the noise information at each step is obtained;

[0068] S3: selecting a number of steps of the denoising process as selected steps according to a preset method;

[0069] S4: according to the image difference of each step after removing the noise information, respectively calculating the accumulated image difference between every two adjacent selected steps, and respectively generating a first intermediate mask image corresponding to each accumulated image difference, and a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0070] S5: superimposing the noise image onto any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and updating the superimposed original first intermediate mask image to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0071] S6: performing a pixel-by-pixel binarization process on each of the second intermediate mask images, using the binarized image as a final mask image, and a plurality of the final mask images constitute the multiple exposure mask image set.

[0072] In this embodiment, in the above step S1, the initial mask image is usually a binary image, which contains all the pattern information that needs to be photolithographically onto the chip, that is, which areas need to be exposed (usually white or bright areas) and which areas need to remain unexposed (usually black or dark areas).

[0073] In the above step S2, the training of the above diffusion model can be carried out by the model training method in the prior art. Exemplarily, the following steps are included: Preparation of training data set: The data set in this embodiment is mainly composed of mask images, and no annotation is required; the data in the training data set is divided into a training set and a validation set according to a preset ratio; an initial diffusion model is selected, such as a deep generative network (such as a U-Net structure), and the initial value of the diffusion model is set. The designed diffusion model can predict the mean and variance of the image in the process of adding noise and denoising, and make the intermediate process images of the process of adding noise and denoising satisfy the normal distribution; the mask image in the training set is input into the initial diffusion model, the initial parameters are set, and the training process is carried out in a supervised learning manner. During the training process, the maximum likelihood function can be used as the loss function L(θ):

[0074]

[0075] To obtain a diffusion model to be verified; input the verification data in the verification set into the diffusion model to be verified for verification processing to obtain a verification result; determine whether the verification result is passed; if so, the diffusion model to be verified is used as the diffusion model of an embodiment of the present invention.

[0076] The noise addition process gradually adds noise information to the initial mask image. After each step of adding noise, a more blurred image, i.e., a noisy image, is obtained. The denoising process is used to remove noise from the noisy image and restore a clear image. After each step of denoising, an intermediate image closer to the original image is obtained, and the noise information removed at each step is recorded, i.e., the above-mentioned image difference. The image difference is the difference value vector between the two images, which can be obtained by calculating the difference in the pixel value of each pixel. The pixel value can be, for example, the brightness or grayscale of each pixel.

[0077] In the above step S3, a specific number of steps may be selected as the selected number of steps from all the number of steps of the denoising process based on experience or historical experimental data.

[0078] In the above step S4, for each pair of adjacent selected steps, the accumulated image difference between them is calculated. The accumulated image difference is the change in the image during the denoising process from one step to the next. Based on these accumulated image differences, the corresponding first intermediate mask image is generated.

[0079] In the above step S5, the noise image of step S2 is superimposed on any image in the first intermediate mask image set to generate a new superimposed intermediate mask image. The original image is replaced with the new superimposed image to update the intermediate mask image set. The sum of the second intermediate mask images in the intermediate mask image set is the noise image at the starting point of the denoising process plus the image change amount at each step in the entire denoising process, that is, the sum of the second intermediate mask images is the target image. In this way, the image is automatically segmented through the denoising step of the diffusion model without manual segmentation.

[0080] In the above step S6, the mask image for multiple exposures should be a binary image. However, the image obtained in the above step S5 is not a binary image, so it is necessary to perform a binarization process on each pixel, and finally obtain a set of final mask images composed entirely of binary images, which can be directly used in the multiple exposure lithography process. The binarization process is to convert the value of each pixel into 0 or 1 (or black or white).

[0081] Through the above steps, this embodiment realizes automatic segmentation of the initial mask image and obtains a set of multi-exposure mask images without relying on manual experience for segmentation, which greatly improves efficiency and provides reliable and stable results.

[0082] In a specific embodiment, the step S2 of inputting the initial mask image into a pre-trained diffusion model for denoising further includes:

[0083] S201: removing noise information from the noisy image step by step through the diffusion model, and obtaining information difference of the noise information removed at each step.

[0084] In this embodiment, in the above step S201, in each step of the denoising process, the diffusion model predicts and removes a part of the noise according to the noise situation of the current image. For each step of the denoising process, the information amount of each generated denoised intermediate image can be obtained, and the information amount is the mean or variance of the predefined parameters of the image, and the predefined parameters can be, for example, pixel brightness, grayscale, etc. The above information difference is the information amount change value of each step of the image from the current state to a clearer state during the denoising process. The calculation of the information difference can be achieved by comparing the information amount difference of two adjacent step images. Specifically, the value difference between the image obtained after each step of denoising and the image of the previous step at the corresponding pixel position can be calculated, and then the value difference at all pixel positions is summarized (such as summation, average, variance, etc.) to obtain the information difference of the step. The embodiment of the present invention creatively introduces the calculation of the information amount and information difference of each step in the denoising process, so that in actual use, the above information amount and information difference can be used to guide the generation strategy of the multi-exposure mask image set of different initial mask images.

[0085] In a specific embodiment, the step S3 of selecting a number of steps of the denoising process as the selected steps in a preset manner includes:

[0086] S301: Setting the intermediate information difference threshold;

[0087] S302: Determine a number of selected steps from the steps of the denoising process so that the cumulative information difference between adjacent selected steps is within the intermediate information difference threshold; wherein the cumulative information difference between adjacent selected steps is obtained by superimposing the information difference of each step between adjacent selected steps after removing the noise information.

[0088] In this embodiment, the intermediate information difference threshold is used to control the cumulative information difference between adjacent selected steps to ensure that the image denoising process is neither too sharp (resulting in loss of image details) nor too slow (resulting in low computational efficiency) under the selected number of steps. The setting of the intermediate information difference threshold can be determined based on experience, experimental data or actual needs. In a specific embodiment, for example, the number of noise adding steps is 700, and the information difference of removing noise information in each step is recorded as dx t , set an initial selected step number t0, for example, t0 = 1, then after the denoising process of the i-th step, if the superimposed cumulative information difference is within the intermediate information difference threshold, then the i-th step is taken as a selected step number. Then, according to the above method, calculate the next selected step number starting from the i-th step. In this embodiment, for different initial mask images, the specific intermediate information difference threshold can be adjusted so that the final multi-exposure mask image set meets the requirements of chip manufacturing and achieves the desired accuracy and yield.

[0089] In a specific embodiment, the step S6 of performing pixel-by-pixel binarization on each of the second intermediate mask images and using the binarized images as the final mask images includes:

[0090] S601: Compare the pixel values ​​of the first designated pixels at the same position in each of the second intermediate mask images to obtain a first maximum pixel value;

[0091] S602: setting the pixel value of the first designated pixel in the second intermediate mask image corresponding to the first maximum pixel value to 1, and setting the pixel values ​​of the first designated pixels in the remaining second intermediate mask images to 0;

[0092] S603: taking each image after traversing all pixels as the final mask image.

[0093] In this embodiment, the above pixel value may be the brightness or grayscale corresponding to the pixel.

[0094] In step S602, according to the pixel position corresponding to the first maximum pixel value obtained in step S601, the pixel value of the first designated pixel in the image is set to 1, indicating that the position needs to be exposed in the multiple exposure lithography process. At the same time, the pixel values ​​of the same positions in all other second intermediate mask images are set to 0, indicating that these positions do not need to be exposed in the multiple exposure process. The process of steps S601 and S602 is repeated until all pixels in the second intermediate mask image are traversed, and the final mask image set after pixel-by-pixel binarization is obtained. Each pixel in these images has been set to 0 or 1 and can be directly used in the multiple exposure lithography process.

[0095] In a specific embodiment, the step S603 of using each image after traversing all pixels as the final mask image includes:

[0096] S6031: Obtaining an image difference between the target image and the initial mask image;

[0097] S6032: Obtain the position of the difference pixel point corresponding to the image difference;

[0098] S6033: Setting the pixel values ​​of non-zero pixel points of the difference pixel points in the final mask image to 0 to obtain an updated final mask image, and using the updated final mask image to replace the original image in the multiple exposure mask image set.

[0099] In this embodiment, the image obtained after the initial mask image is processed by adding noise and reducing noise through the diffusion model may have slight differences, so the mask image needs to be adjusted.

[0100] In step S6031, the image difference is a set of pixel vectors with differences, and the differences may be differences in attributes such as brightness. Since the initial mask image is a binary image, for each pixel difference comparison, if there is a brightness difference, it means that there is an image difference at that point.

[0101] In step S6032, the positions of the difference pixels are determined, which are the positions that need to be adjusted.

[0102] In step S6033, for each non-zero pixel located at the position of the difference pixel in each image, its pixel value is set to 0, so as to correct the final mask image so that it accurately matches the features of the initial mask image.

[0103] In a specific embodiment, the step S6 of performing pixel-by-pixel binarization on each of the second intermediate mask images and using the binarized images as the final mask images includes:

[0104] S611: Acquire a pixel position corresponding to an exposed lithography image in the initial mask image, and record each pixel corresponding to the exposed lithography image as a second designated pixel;

[0105] S612: setting the pixel values ​​of the pixels in each of the second intermediate mask images that are at different positions from the second designated pixel to 0;

[0106] S613: Compare the pixel values ​​at the same pixel position as the second designated pixel in each of the second intermediate mask images to obtain a second maximum pixel value;

[0107] S614: setting the pixel value of the pixel position corresponding to the second maximum pixel value in the second intermediate mask image corresponding to the second maximum pixel value to 1; and setting the pixel values ​​of the pixel positions corresponding to the second maximum pixel value in the remaining second intermediate mask images to 0;

[0108] S615: taking each image after traversing all pixels as the final mask image.

[0109] This embodiment provides another method for correcting a mask image.

[0110] In step S611, since the initial mask image is a binary image, the position corresponding to the exposed photolithography image is the position with a pixel value of 1, therefore, all pixel positions with a pixel value of 1 in the initial mask image can be extracted to obtain the exposed photolithography image.

[0111] In step S612, for each second mask image, the pixel values ​​of the pixels not matching the second designated pixel position (ie, the pixels in the non-exposure lithography image area) are all set to 0, so that the subsequent steps only calculate the pixels related to the exposure lithography.

[0112] In steps S613-S6015, the process is similar to the aforementioned steps S601-S603, and the pixel values ​​of the pixel positions corresponding to the exposed lithography images in each second mask image are binarized to obtain a final mask image set after pixel-by-pixel binarization. Each pixel in these images has been set to 0 or 1, and can be directly used in the multiple exposure lithography process.

[0113] In a specific embodiment, the step S3 of selecting a number of steps of the denoising process as the selected steps in a preset manner includes:

[0114] S311: Setting the starting number of steps and the step interval;

[0115] S312: superimposing the step interval on the starting step number, and taking the number of steps obtained by superimposing the step interval each time as the selected step number, until the superimposed value exceeds the maximum number of steps for denoising processing.

[0116] In this embodiment, illustratively, for a denoising process that undergoes 700 steps, the starting step number is set to 1, the step interval is set to 100, and the selected steps are 100, 200, 300, 400, 500, and 600.

[0117] In another embodiment, multiple step intervals can be set. For example, for a denoising process that undergoes 700 steps, the starting step number is set to 1, the step intervals are set to 49, 149, 199, 300, and finally the selected steps are set to 49, 199, 399, 700.

[0118] like Figure 3 As shown, in one embodiment, a device for generating a multi-exposure mask image set is provided. The device comprises:

[0119] An acquisition module 10, used for acquiring an initial mask image;

[0120] The noise processing module 20 is used to input the initial mask image into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0121] A selection module 30, configured to select a number of steps of the denoising process as selected steps according to a preset method;

[0122] A first image set generating module 40 is used to calculate the accumulated image difference between every two adjacent selected steps according to the image difference of each step of removing noise information, and to generate a first intermediate mask image corresponding to each accumulated image difference, wherein a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0123] The second image set generating module 50 is used to superimpose the noise image onto any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and update the superimposed original first intermediate mask image to the superimposed intermediate mask image, so that the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0124] The mask image set generating module 60 is used to perform pixel-by-pixel binarization on each of the second intermediate mask images, and use the binarized image as a final mask image. Several of the final mask images constitute the multiple exposure mask image set.

[0125] In a specific embodiment, the noise processing module 20 further includes:

[0126] The information difference acquisition unit is used to gradually remove noise information from the noise image through the diffusion model, and obtain the information difference of each step of removing the noise information.

[0127] In a specific embodiment, the selection module 30 includes:

[0128] A first setting unit, used to set an intermediate information difference threshold;

[0129] The first selected step determination unit is used to determine a number of selected steps from the steps of the denoising process so that the cumulative information difference between adjacent selected steps is within the intermediate information difference threshold; wherein the cumulative information difference between adjacent selected steps is obtained by superimposing the information difference of each step between adjacent selected steps after removing the noise information.

[0130] In a specific embodiment, the mask image set generating module 60 includes:

[0131] A first comparison unit, configured to compare pixel values ​​of first designated pixels at the same position in each of the second intermediate mask images to obtain a first maximum pixel value;

[0132] a first binarization unit, configured to set the pixel value of the first designated pixel in the second intermediate mask image corresponding to the first maximum pixel value to 1, and set the pixel values ​​of the first designated pixels in the remaining second intermediate mask images to 0;

[0133] The first traversal unit is used to use each image after traversing all pixels as the final mask image.

[0134] In a specific embodiment, the first traversal unit includes:

[0135] A difference acquisition subunit, used for acquiring an image difference between the target image and the initial mask image;

[0136] A pixel position acquisition subunit, used to acquire the position of a difference pixel point corresponding to the image difference;

[0137] The first binarization subunit is used to set the pixel values ​​of non-zero pixel points of the difference pixel points in the final mask image to 0 to obtain an updated final mask image, and use the updated final mask image to replace the original image in the multiple exposure mask image set.

[0138] In a specific embodiment, the mask image set generating module 60 includes:

[0139] A pixel position acquisition unit, used for acquiring a pixel position corresponding to the exposure photolithography image in the initial mask image, and recording each pixel corresponding to the exposure photolithography image as a second designated pixel;

[0140] a revision unit, configured to set the pixel value of the pixel in each of the second intermediate mask images, which pixel position is different from the second designated pixel, to 0;

[0141] A second comparison unit, configured to compare pixel values ​​at the same pixel position as the second designated pixel in each of the second intermediate mask images to obtain a second maximum pixel value;

[0142] a second binarization unit, configured to set the pixel value of the pixel position corresponding to the second maximum pixel value in the second intermediate mask image corresponding to the second maximum pixel value to 1; and set the pixel values ​​of the pixel positions corresponding to the second maximum pixel value in the remaining second intermediate mask images to 0;

[0143] The second traversal unit is used to use each image after traversing all pixels as a final mask image.

[0144] In a specific embodiment, the selection module 30 includes:

[0145] A second setting unit is used to set the starting number of steps and the step interval;

[0146] The second selected step number determination unit is used to superimpose the step number interval on the starting step number, and use the number of steps obtained by superimposing the step number interval each time as the selected step number, until the superimposed value exceeds the maximum number of steps for denoising processing.

[0147] This embodiment realizes automatic segmentation of the initial mask image to obtain a set of multi-exposure mask images without relying on manual experience for segmentation, which greatly improves efficiency and provides reliable and stable results.

[0148] Figure 4 FIG. 1 shows an internal structure diagram of a computer device in an embodiment. The computer device may be a terminal or a server. Figure 4As shown, the computer device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor may implement the method for generating a multiple exposure mask image set. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor may implement the method for generating a multiple exposure mask image set. Those skilled in the art will appreciate that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0149] In one embodiment, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps:

[0150] Obtain an initial mask image;

[0151] The initial mask image is input into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0152] Selecting a number of steps of the denoising process as selected steps according to a preset method;

[0153] According to the image difference of each step of removing noise information, the accumulated image difference between every two adjacent selected steps is calculated respectively, and a first intermediate mask image corresponding to each accumulated image difference is generated respectively, and a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0154] The noise image is superimposed on any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and the superimposed original first intermediate mask image is updated to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0155] Each of the second intermediate mask images is subjected to a pixel-by-pixel binarization process, and the binarized image is used as a final mask image. A plurality of the final mask images constitute the multiple exposure mask image set.

[0156] This embodiment realizes automatic segmentation of the initial mask image to obtain a set of multi-exposure mask images without relying on manual experience for segmentation, which greatly improves efficiency and provides reliable and stable results.

[0157] In one embodiment, a computer-readable storage medium is provided, storing a computer program, wherein when the computer program is executed by a processor, the processor performs the following steps:

[0158] Obtain an initial mask image;

[0159] The initial mask image is input into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained.

[0160] Selecting a number of steps of the denoising process as selected steps according to a preset method;

[0161] According to the image difference of each step of removing noise information, the accumulated image difference between every two adjacent selected steps is calculated respectively, and a first intermediate mask image corresponding to each accumulated image difference is generated respectively, and a plurality of the first intermediate mask images constitute a first intermediate mask image set;

[0162] The noise image is superimposed on any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and the superimposed original first intermediate mask image is updated to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image;

[0163] Each of the second intermediate mask images is subjected to a pixel-by-pixel binarization process, and the binarized image is used as a final mask image. A plurality of the final mask images constitute the multiple exposure mask image set.

[0164] This embodiment realizes automatic segmentation of the initial mask image to obtain a set of multi-exposure mask images without relying on manual experience for segmentation, which greatly improves efficiency and provides reliable and stable results.

[0165] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0166] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0167] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for generating a multi-exposure mask image set, characterized in that: The method comprises: Obtain an initial mask image; The initial mask image is input into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained. Selecting a number of steps of the denoising process as selected steps according to a preset method; According to the image difference of each step of removing noise information, the accumulated image difference between every two adjacent selected steps is calculated respectively, and a first intermediate mask image corresponding to each accumulated image difference is generated respectively, and a plurality of the first intermediate mask images constitute a first intermediate mask image set; The noise image is superimposed on any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and the superimposed original first intermediate mask image is updated to the superimposed intermediate mask image, the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image; Each of the second intermediate mask images is subjected to a pixel-by-pixel binarization process, and the binarized image is used as a final mask image. A plurality of the final mask images constitute the multiple exposure mask image set.

2. The method for generating a multi-exposure mask image set according to claim 1, characterized in that: The step of inputting the initial mask image into a pre-trained diffusion model for denoising further includes: The diffusion model is used to gradually remove noise information from the noise image, and the information difference of the noise information removed at each step is obtained.

3. The method for generating a multi-exposure mask image set according to claim 2, characterized in that: The step of selecting a number of steps of the denoising process as selected steps in a preset manner comprises: Set the intermediate information difference threshold; From the steps of the denoising process, a number of selected steps are determined so that the cumulative information difference between adjacent selected steps is within the intermediate information difference threshold; wherein the cumulative information difference between adjacent selected steps is obtained by superimposing the information difference of each step between adjacent selected steps after removing the noise information.

4. The method for generating a multi-exposure mask image set according to claim 1, characterized in that: The step of performing pixel-by-pixel binarization on each of the second intermediate mask images and using the binarized images as the final mask images comprises: Comparing the pixel values ​​of the first designated pixels at the same position in each of the second intermediate mask images to obtain a first maximum pixel value; Setting the pixel value of the first designated pixel in the second intermediate mask image corresponding to the first maximum pixel value to 1, and setting the pixel values ​​of the first designated pixels in the remaining second intermediate mask images to 0; Each image after traversing all pixels is used as the final mask image.

5. The method for generating a multi-exposure mask image set according to claim 4, characterized in that: The step of using each image after traversing all pixels as the final mask image includes: Acquire an image difference between the target image and the initial mask image; Obtaining the position of the difference pixel point corresponding to the image difference; The pixel values ​​of non-zero pixels of the difference pixels in the final mask image are set to 0 to obtain an updated final mask image, and the updated final mask image is used to replace the original image in the multiple exposure mask image set.

6. The method for generating a multi-exposure mask image set according to claim 1, characterized in that: The step of performing pixel-by-pixel binarization on each of the second intermediate mask images and using the binarized images as the final mask images comprises: Acquire a pixel position corresponding to an exposed photolithography image in the initial mask image, and record each pixel corresponding to the exposed photolithography image as a second designated pixel; setting the pixel value of the pixel at a different pixel position from the second designated pixel in each of the second intermediate mask images to 0; Compare pixel values ​​at the same pixel position as the second designated pixel in each of the second intermediate mask images to obtain a second maximum pixel value; Setting the pixel value of the pixel position corresponding to the second maximum pixel value in the second intermediate mask image corresponding to the second maximum pixel value to 1; and setting the pixel values ​​of the pixel positions corresponding to the second maximum pixel value in the remaining second intermediate mask images to 0; Each image after traversing all pixels is used as the final mask image.

7. The method for generating a multi-exposure mask image set according to claim 1, characterized in that: The step of selecting a number of steps of the denoising process as selected steps in a preset manner comprises: Set the starting number of steps and step interval; The step interval is superimposed on the starting step number, and the number of steps obtained by superimposing the step interval each time is used as the selected step number, until the superimposed value exceeds the maximum number of steps for denoising processing.

8. A device for generating a multi-exposure mask image set, characterized in that: The device comprises: An acquisition module, used for acquiring an initial mask image; A noise processing module is used to input the initial mask image into a pre-trained diffusion model for noise addition and denoising. The noise addition is used to gradually add noise information to the initial mask image, and the image obtained after the noise addition is completed is used as the noise image. The denoising is used to gradually remove noise information from the noise image, and the image obtained after the denoising is completed is used as the target image, and the image difference after the noise information is removed at each step is obtained. A selection module, used for selecting a number of steps of the denoising process as selected steps according to a preset method; A first image set generating module, configured to calculate the accumulated image difference between every two adjacent selected steps according to the image difference of each step of removing noise information, and to generate a first intermediate mask image corresponding to each accumulated image difference, wherein a plurality of the first intermediate mask images constitute a first intermediate mask image set; A second image set generating module, configured to superimpose the noise image onto any one of the first intermediate mask images in the first intermediate mask image set to generate a superimposed intermediate mask image, and update the superimposed original first intermediate mask image to the superimposed intermediate mask image, so that the first intermediate mask image set is updated to a second intermediate mask image set, and each mask image in the second intermediate mask image set is used as a second intermediate mask image; The mask image set generation module is used to perform pixel-by-pixel binarization on each of the second intermediate mask images, and use the binarized image as the final mask image. Several of the final mask images constitute the multiple exposure mask image set.

9. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor executes the steps of the method for generating a multiple exposure mask image set according to any one of claims 1 to 7.

10. A computer device, characterized in that: The device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method for generating a multiple-exposure mask image set according to any one of claims 1 to 7.

Images