Method, device, system and medium for contour extraction of noisy electron beam images

Abstract

The application provides a contour extraction method, device, system and medium of a noisy electron beam image, a filtering interval is determined according to a three-sigma principle by using an average value and a standard deviation, and a segmented linear transformation is performed on a filtered electron beam grayscale image; a filtering algorithm is used to filter an enhanced electron beam grayscale image; an edge extraction operator is used to extract edges of the filtered electron beam grayscale image, a primary electron beam grayscale image contour and an original electron beam grayscale image are superimposed, and an iterative operation is performed on the superimposed electron beam grayscale image; a morphological processing is performed on an iterative electron beam grayscale image obtained through iteration; and an image operation is performed to obtain a target electron beam single-edge image. The filtering interval is set, a preliminary noise reduction processing is performed according to the three-sigma principle, image points are protected and part of the noise points are removed, and the noise elimination efficiency is improved. The segmented linear transformation is performed, the image contrast is enhanced, and the contour extraction is more accurate. The morphological processing is performed, and a precise single-edge image is obtained through the operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a contour extraction method, device, system and medium for a noisy electron beam image. BACKGROUND

[0002] With the continuous development of integrated circuit manufacturing technology, the pattern critical dimension is getting smaller and smaller, and the industry has not relaxed the requirements for pattern edge quality.

[0003] For the edge part of the image, it is not allowed to exist photoresist residue, structure adhesion, bridging and other abnormal situations. In the semiconductor production process, when the line width size is smaller than the wavelength of the lithography light source, the image will be distorted due to the diffraction effect of light, so that the imaging on the silicon wafer is different from the design layout, and the optical proximity effect such as edge roundness, uneven line width and pattern distortion occurs. The optical proximity effect will affect the appearance and electrical performance of the product, and reduce the yield of the chip.

[0004] In order to cope with the influence caused by the diffraction effect of light, the industry has proposed optical proximity correction (OPC) technology. The optical proximity correction technology is a kind of lithography resolution enhancement technology, which uses a calculation method to correct the pattern on the mask, adjusts the topological structure of the light transmission area pattern on the mask or adds a small auxiliary pattern on the mask, so that the photoresist imaging result is as close as possible to the mask pattern. The technology can be used to correct the diffraction effect to ensure the electrical performance of the production circuit.

[0005] The optical proximity correction technology is divided into empirical-based optical proximity correction and model-based optical proximity correction. The empirical-based optical proximity correction technology is greatly influenced by experience and pattern complexity, and has low precision; the model-based optical proximity correction technology calculates the pattern after exposure using optical model and photoresist chemical model, which needs a large amount of data for fitting.

[0006] In order to ensure the correction accuracy of the optical proximity effect correction mask, a high-precision optical model and a photoresist model need to be used, especially the establishment of the photoresist model, which needs to use a large number of test patterns for modeling. The more data, the more accurate, the higher the modeling accuracy. In the actual modeling process, the test pattern used in the traditional modeling is mostly one-dimensional pattern, and more key size data is used for modeling. However, when applied to two-dimensional patterns, due to the increase of pattern complexity, the model accuracy based on key size measurement data will be greatly limited and restricted, so that the optical proximity effect correction accuracy and the correction pattern quality often deviate from the results of the exposed patterns. Therefore, when establishing the two-dimensional pattern optical proximity effect correction model (photoresist model), the pattern contour is selected for modeling. The edge of the two-dimensional image is the place where the regional attribute changes suddenly, the brightness changes greatly at the image edge, and the image information is the most concentrated place, which contains rich information. Using pattern contour modeling can help to enrich the types of modeling patterns, improve the applicability and accuracy of the model to complex patterns.

[0007] However, as the technology node shrinks, the decrease of the layout size leads to the increase of the noise of the electron beam measurement image and the decrease of the signal-to-noise ratio, that is, the obtained two-dimensional pattern has very obvious image noise, which greatly restricts the accurate extraction of the contour pattern and further the optical proximity effect correction modeling.

[0008] Affected by the image noise, when the contour is extracted, there are problems such as pseudo-edge, multi-edge, or edge missing and breaking, and the extraction effect is not good. Therefore, how to improve the accuracy of the contour extraction of the noisy electron beam image and improve the extraction effect is a technical problem to be solved in the field. SUMMARY

[0009] Therefore, this summary part is provided to briefly introduce the concepts, which will be described in detail in the following specific embodiments part. This summary part is not intended to identify key or essential features of the claimed technical solutions or to limit the scope of the claimed technical solutions.

[0010] The purpose of the present application is to provide a contour extraction method, device, system and medium for a noisy electron beam image, which can improve the accuracy of the contour extraction of the noisy electron beam image and improve the extraction effect.

[0011] To achieve the above-mentioned purpose, the present application has the following technical solutions:

[0012] In a first aspect, the present application provides a contour extraction method for a noisy electron beam image, comprising:

[0013] obtaining an original electron beam gray-scale image; calculating the average value and the standard deviation of the gray-scale value of the pixel points of the original electron beam gray-scale image;

[0014] determine a filtering interval according to a three-sigma principle using the average value and the standard deviation, filter the original electron beam gray scale image according to the filtering interval to obtain a filtered electron beam gray scale image;

[0015] perform piecewise linear transformation on the filtered electron beam gray scale image to obtain an enhanced electron beam gray scale image, filter the enhanced electron beam gray scale image using an adaptive median filtering algorithm to obtain a filtered electron beam gray scale image, and perform edge extraction on the filtered electron beam gray scale image using an edge extraction operator to obtain a primary electron beam gray scale image contour;

[0016] superimpose the primary electron beam gray scale image contour and the original electron beam gray scale image to obtain a superimposed electron beam gray scale image, and perform denoising and edge extraction iteration on the superimposed electron beam gray scale image;

[0017] when the iteration reaches a preset number of times, perform morphological processing on an iteration electron beam gray scale image obtained through iteration to obtain a target electron beam image edge contour curve image;

[0018] perform image and operation on the target electron beam image edge contour curve image to obtain a target electron beam single-edge image.

[0019] In a possible implementation, the morphological processing on the iteration electron beam gray scale image to obtain the target electron beam image edge contour curve image comprises:

[0020] performing closed operation on the iteration electron beam gray scale image to obtain an operation electron beam gray scale image;

[0021] performing hole filling on the first electron beam gray scale image to obtain a filled electron beam gray scale image;

[0022] marking connected domains in the filled electron beam gray scale image, and removing the connected domains with an area smaller than a preset value to obtain the target electron beam image edge contour curve image.

[0023] In a possible implementation, the obtaining of the original electron beam gray scale image comprises:

[0024] acquiring a current electron beam image;

[0025] determining whether the current electron beam image is a single-channel gray scale image;

[0026] if yes, taking the current electron beam image as the original electron beam gray scale image;

[0027] if no, performing gray scale conversion on the current electron beam image to obtain the original electron beam gray scale image.

[0028] In a possible implementation, the denoising and edge extraction iteration on the superimposed electron beam grayscale image comprises:

[0029] Denoising the superimposed electron beam grayscale image by using an adaptive median filtering algorithm to obtain a denoised electron beam grayscale image;

[0030] Edge extraction is performed on the denoised electron beam grayscale image by using an edge extraction operator to obtain an edge extracted electron beam grayscale image.

[0031] In a second aspect, the embodiments of the present application provide a contour extraction device for a noisy electron beam image, comprising:

[0032] An acquisition unit is configured to acquire an original electron beam grayscale image, and calculate an average value and a standard deviation of pixel grayscale values of the original electron beam grayscale image.

[0033] A filtering unit is configured to determine a filtering interval according to a three-sigma principle by using the average value and the standard deviation, filter the original electron beam grayscale image according to the filtering interval, and obtain a filtered electron beam grayscale image.

[0034] A contour unit is configured to perform piecewise linear transformation on the filtered electron beam grayscale image to obtain an enhanced electron beam grayscale image, filter the enhanced electron beam grayscale image by using an adaptive median filtering algorithm to obtain a filtered electron beam grayscale image, and perform edge extraction on the filtered electron beam grayscale image by using an edge extraction operator to obtain a primary electron beam grayscale image contour.

[0035] A superimposition unit is configured to superimpose the primary electron beam grayscale image contour and the original electron beam grayscale image to obtain a superimposed electron beam grayscale image, and perform denoising and edge extraction iteration on the superimposed electron beam grayscale image.

[0036] A morphology unit is configured to, when the iteration reaches a preset number of times, perform morphological processing on an iteration electron beam grayscale image obtained by iteration to obtain a target electron beam image edge contour curve image.

[0037] An operation unit is configured to perform image and operation on the target electron beam image edge contour curve image to obtain a target electron beam single edge image.

[0038] In a possible implementation, the morphology unit is specifically configured to:

[0039] Perform a closing operation on the iteration electron beam grayscale image to obtain an operation electron beam grayscale image.

[0040] Perform hole filling on the first electron beam grayscale image to obtain a filled electron beam grayscale image.

[0041] Labeling connected domains in the filled electron beam gray scale image, removing the connected domains with an area less than a preset value to obtain the target electron beam image edge contour curve image.

[0042] In a possible implementation, the acquisition unit is specifically configured to:

[0043] acquire a current electron beam image;

[0044] determine whether the current electron beam image is a single-channel gray scale image;

[0045] if yes, take the current electron beam image as the original electron beam gray scale image;

[0046] if no, perform gray scale conversion on the current electron beam image to obtain the original electron beam gray scale image.

[0047] In a possible implementation, the superimposition unit is specifically configured to:

[0048] perform denoising on the superimposed electron beam gray scale image by using an adaptive median filtering algorithm to obtain a denoised electron beam gray scale image;

[0049] perform edge extraction on the denoised electron beam gray scale image by using an edge extraction operator to obtain an edge extracted electron beam gray scale image.

[0050] In a third aspect, an embodiment of the present application provides a contour extraction system for a noisy electron beam image, including:

[0051] a memory configured to store a computer program;

[0052] a processor configured to execute the computer program to implement the steps of the contour extraction method for the noisy electron beam image.

[0053] In a fourth aspect, an embodiment of the present application provides a computer readable medium, and the computer readable medium stores a computer program, and the computer program is executed by a processor to implement the steps of the contour extraction method for the noisy electron beam image.

[0054] Compared with the prior art, the embodiment of the present application has the following beneficial effects:

[0055] The embodiment of the application provides a contour extraction method, device, system and medium of a noisy electron beam image, the method comprises the following steps: obtaining an original electron beam gray image; calculating the average value and the standard deviation of the gray value of the pixel points of the original electron beam gray image; determining a filtering interval according to the three-sigma principle by using the average value and the standard deviation, filtering the original electron beam gray image according to the filtering interval, and obtaining a filtered electron beam gray image; performing piecewise linear transformation on the filtered electron beam gray image, and obtaining an enhanced electron beam gray image; filtering the enhanced electron beam gray image by using an adaptive median filtering algorithm, and obtaining a filtered electron beam gray image; performing edge extraction on the filtered electron beam gray image by using an edge extraction operator, and obtaining a primary electron beam gray image contour; superimposing the primary electron beam gray image contour and the original electron beam gray image, obtaining a superimposed electron beam gray image, and performing denoising and edge extraction iteration on the superimposed electron beam gray image; when the iteration reaches a preset number of times, performing morphological processing on the iteration electron beam gray image obtained through iteration, and obtaining a target electron beam image edge contour curve image; and performing image and operation on the target electron beam image edge contour curve image, and obtaining a target electron beam single-edge image. Therefore, the method for setting the filtering interval is adopted, preliminary noise reduction processing is performed according to the three-sigma principle, part of the noise points is removed while the image points are protected, and the efficiency of eliminating the noise is improved. The partial linear change is performed, the contrast of the image is enhanced, and the contour extraction is more accurate. The morphological processing is combined, the accurate single-edge image is obtained by using the operation, the accuracy of the contour extraction of the noisy electron beam image is improved, and the extraction effect is improved. BRIEF DESCRIPTION OF DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0057] The above and other features, advantages, and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the original and elements are not necessarily drawn according to the scale.

[0058] Figure 1 A flowchart of a contour extraction method of a noisy electron beam image provided by the embodiment of the present application is shown;

[0059] Figure 2 A schematic diagram of an original electron beam gray image provided by the embodiment of the present application is shown;

[0060] Figure 3 A gray scale distribution diagram of the original electron beam gray scale image is shown;

[0061] Figure 4 A schematic diagram of the initial electron beam gray scale image profile is shown;

[0062] Figure 5 A schematic diagram of the iterative electron beam gray scale image is shown;

[0063] Figure 6 A schematic diagram of the target electron beam image edge profile curve image is shown;

[0064] Figure 7 A schematic diagram of the target electron beam single edge image is shown;

[0065] Figure 8 A schematic diagram of the profile extraction device of the electron beam image containing noise is shown. DETAILED DESCRIPTION

[0066] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0067] In the following description, a lot of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the spirit of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0068] As described in the background, with the continuous development of integrated circuit manufacturing technology, the pattern critical dimension is getting smaller and smaller, and the industry has not relaxed the requirements for pattern edge quality.

[0069] For the edge part of the image, it is not allowed to exist photoresist residue, structure adhesion, bridging and other abnormal situations. In the semiconductor production process, when the line width size is less than the wavelength of the lithography light source, due to the diffraction effect of light, the image will be distorted, so that the imaging on the silicon wafer is different from the design layout, and the optical proximity effect such as edge roundness, line width unevenness, pattern distortion occurs. Optical proximity effect will affect the appearance and electrical performance of the product, and reduce the yield of the chip.

[0070] In order to cope with the influence caused by the diffraction effect of light, the industry proposes an optical proximity correction (OPC) technology. The optical proximity correction technology is a kind of lithography resolution enhancement technology, which uses a calculation method to correct the pattern on the mask, adjusts the topological structure of the light transmission area pattern on the lithography mask or adds a small auxiliary pattern on the mask, so that the photoresist imaging result is as close as possible to the mask pattern. The technology can be used to correct the diffraction effect to ensure the electrical performance of the produced circuit.

[0071] The optical proximity correction technology is divided into empirical-based optical proximity correction and model-based optical proximity correction. The empirical-based optical proximity correction technology is greatly affected by experience and pattern complexity, and has low precision; the model-based optical proximity correction technology calculates the exposed pattern using an optical model and a photoresist chemical model, and a large amount of data is needed for fitting.

[0072] In order to ensure the correction accuracy of the optical proximity correction mask, a high-precision optical model and a photoresist model need to be used, especially the establishment of the photoresist model, which needs to use a large amount of test patterns for modeling. The more data, the more accurate, and the higher the modeling accuracy. In the actual modeling process, the test layout used in the traditional modeling is mostly one-dimensional pattern, and more key size data is used for modeling. However, when applied to two-dimensional patterns, due to the increase of pattern complexity, the model accuracy based on key size measurement data will be greatly limited and restricted, so that the optical proximity correction accuracy and the correction layout quality often deviate from the result of the exposed pattern. Therefore, when establishing the two-dimensional pattern optical proximity correction model (photoresist model), the contour of the pattern is selected for modeling. The edge of the two-dimensional image is the place where the region attribute changes suddenly, and the brightness changes greatly at the image edge, which is the most concentrated place of image information and contains rich information. Using contour modeling can help to enrich the types of modeling patterns and improve the applicability and accuracy of the model to complex patterns.

[0073] However, as the technology node shrinks, the decrease of layout size leads to the increase of noise of electron beam measured image and the decrease of signal-to-noise ratio, that is, the obtained two-dimensional pattern has very obvious image noise, which greatly restricts the accurate extraction of contour pattern and further optical proximity correction modeling.

[0074] Affected by image noise, false edges, multiple edges, or edge missing and breaking may occur when extracting the contour, and the extraction effect is not good. Therefore, how to improve the accuracy of contour extraction of noisy electron beam image and improve the extraction effect is a technical problem to be solved in the field.

[0075] To solve the above technical problems, the embodiment of the present application provides a contour extraction method, device, system and medium of a noisy electron beam image, which comprises the following steps: obtaining an original electron beam gray image; calculating the average value and the standard deviation of the gray value of the pixel points of the original electron beam gray image; determining a filtering interval according to the three-sigma principle by using the average value and the standard deviation, filtering the original electron beam gray image according to the filtering interval, and obtaining a filtered electron beam gray image; performing piecewise linear transformation on the filtered electron beam gray image, and obtaining an enhanced electron beam gray image; filtering the enhanced electron beam gray image by using an adaptive median filtering algorithm, and obtaining a filtered electron beam gray image; performing edge extraction on the filtered electron beam gray image by using an edge extraction operator, and obtaining a primary electron beam gray image contour; superimposing the primary electron beam gray image contour and the original electron beam gray image, obtaining a superimposed electron beam gray image, and performing denoising and edge extraction iteration on the superimposed electron beam gray image; when the iteration reaches a preset number of times, performing morphological processing on the iteration electron beam gray image obtained by iteration, and obtaining a target electron beam image edge contour curve image; and performing image and operation on the target electron beam image edge contour curve image, and obtaining a target electron beam single-edge image. Therefore, the method for setting the filtering interval is adopted, the preliminary denoising processing is performed according to the three-sigma principle, part of the noise points is removed while the image points are protected, and the efficiency of eliminating the noise is improved. The partial linear change is performed, the contrast of the image is enhanced, and the contour extraction is more accurate. The morphological processing is combined, the accurate single-edge image is obtained by using the and operation, the accuracy of the contour extraction of the noisy electron beam image is improved, and the extraction effect is improved.

[0076] Exemplary method

[0077] Referring to Figure 1 As shown in FIG. 1, a flowchart of a contour extraction method of a noisy electron beam image provided by the embodiment of the present application comprises the following steps:

[0078] S101: obtaining an original electron beam gray image; and calculating the average value and the standard deviation of the gray value of the pixel points of the original electron beam gray image.

[0079] In the embodiment of the present application, the original electron beam gray image can be obtained; the average value and the standard deviation of the gray value of the pixel points of the original electron beam gray image are calculated.

[0080] Specifically, according to the SEM (scanning electron microscope) image characteristics, the electron beam intensity reflects the surface topography characteristics of the target structure, the output image is represented by using the gray pixel value, and the gray range is 0-255 pixel values. Referring to Figure 2As shown in the figure, it is a schematic diagram of an original electron beam gray scale image provided by an embodiment of the present application. Optionally, in some cases, a CD-SEM (critical dimension SEM) is used for dedicated size measurement, and the image obtained thereby is also a source of the original electron beam gray scale image.

[0081] During the acquisition of the SEM image and the transmission of the image, noise inevitably occurs. The occurrence of image noise affects the accuracy of the profile extraction. Image noise is an important factor that interferes with the correct understanding of image information by humans. Preliminary filtering of image noise can improve the clarity of the profile extraction. Image noise is random, and a probability density function can be selected to describe and analyze the noise. The average value of all pixel gray values of the original gray scale value image is calculated by performing pixel value statistics on the original gray scale value image, and the standard deviation of the pixel gray values is calculated. The average value represents the average level of the image gray scale, and the standard deviation represents the dispersion degree of the image pixel values from the average value.

[0082] Referring to Figure 3 As shown in the figure, it is a gray scale distribution statistical diagram of an original electron beam gray scale image provided by an embodiment of the present application, the horizontal coordinate is the pixel value of the original electron beam gray scale image, and the vertical coordinate is the gray scale value corresponding to the pixel value of the original electron beam gray scale image.

[0083] The figure shows that there are two pixel value intervals and corresponding peak values, wherein the interval with a higher pixel value distribution is the electron beam intensity distribution generated at the top and edge positions of the line structure, and the interval with a lower pixel value distribution is the electron beam intensity distribution generated at the positions between the line structures. In this embodiment, according to the target needs, the higher electron beam intensity distribution interval is selected as the analysis and reference interval.

[0084] S102: determining a filtering interval according to a three-sigma principle by using the average value and the standard deviation, filtering the original electron beam gray scale image according to the filtering interval, and obtaining a filtered electron beam gray scale image.

[0085] In the embodiment of the present application, the filtering interval refers to setting an image point filtering interval. The pixel points within the interval range are determined as image points and are retained, and the pixel points not within the interval range are eliminated.

[0086] In the embodiment of the present application, the mean μ and the standard deviation σ of the original electron beam gray scale image can be calculated. According to the three-sigma principle, the filter interval is preliminarily set as (μ-3σ, μ+3σ). After setting the filter interval, the image histogram is analyzed. The high electron beam intensity distribution interval meets the requirement of the filter interval. Most of the pixel points are concentrated in the confidence interval. By setting the filter interval, the points not in the filter interval range are set as noise points. The pixel points in the filter interval are determined as image points. The purpose of preliminarily eliminating part of the sharp noise is achieved. The range of noise points is reduced. The efficiency of eliminating noise can be improved.

[0087] S103: performing piecewise linear transformation on the filtered electron beam gray scale image to obtain an enhanced electron beam gray scale image; performing filtering on the enhanced electron beam gray scale image by using an adaptive median filtering algorithm to obtain a filtered electron beam gray scale image; and performing edge extraction on the filtered electron beam gray scale image by using an edge extraction operator to obtain a primary electron beam gray scale image contour.

[0088] In the embodiment of the present application, the piecewise linear transformation is a method of image enhancement. The gray scale value is selectively expanded. The image contrast can be enhanced. The visual effect of the image can be improved.

[0089] Specifically, the filtered electron beam gray scale image obtained above can be retained for the interval corresponding to the line structure with low pixel value distribution without processing. The gray scale value of the interval corresponding to the top and edge position of the line structure with high pixel value distribution is linearly mapped to the range of 0-255. After the piecewise linear processing, the feature edge to be extracted can be more prominently displayed.

[0090] Then, the adaptive median filtering algorithm can be used to filter the enhanced electron beam gray scale image to obtain a filtered electron beam gray scale image.

[0091] The denoising processing refers to filtering out the noise in the SEM image, restoring the noise-free image, and retaining the original detail information in the image to the maximum extent. Common denoising algorithms include median filtering, Gaussian filtering, mean filtering, and Wiener filtering.

[0092] Specifically, the adaptive median filtering can be selected to denoise the enhanced electron beam gray scale image. The adaptive median filtering can automatically adjust the size of the filtering window according to the concentration of the noise in the neighborhood, and enhance the performance of removing high-concentration noise. After the adaptive median filtering denoising, the noise points are significantly reduced, and the image edge is clearer.

[0093] The de-noising processing of the embodiment of the present application refers to processing the enhanced electron beam gray scale image, filtering out the noise in the image, restoring the noise-free image, and retaining the original detail information in the image to the maximum extent. Optionally, median filtering, Gaussian filtering, mean filtering, Wiener filtering and other algorithms can be used.

[0094] In addition, referring to Figure 4 Fig. 1 shows a schematic diagram of a primary electron beam gray scale image profile provided by the embodiment of the present application. The embodiment of the present application can also use an edge extraction operator to perform edge extraction on the filtered electron beam gray scale image to obtain the primary electron beam gray scale image profile.

[0095] Specifically, the primary edge extraction refers to using an edge extraction operator to perform edge extraction to obtain the primary electron beam gray scale image profile. The image has the characteristic that the extracted edge is similar to the actual edge in the actual image, and can extract the boundary of the region with large gray scale value change. The obtained is a double-edge image.

[0096] The double-edge image obtained in the embodiment contains real edges and pseudo edges. The pseudo edge refers to a place with large gray scale change in a defect point or background being identified as an edge when using an edge detection operator to perform detection. When performing edge detection, the appearance of the pseudo edge affects the accuracy of image edge extraction.

[0097] S104: superimpose the primary electron beam gray scale image profile and the original electron beam gray scale image to obtain a superimposed electron beam gray scale image, and perform de-noising and edge extraction iteration on the superimposed electron beam gray scale image.

[0098] In the embodiment of the present application, the primary electron beam gray scale image profile and the original electron beam gray scale image can be superimposed to obtain a superimposed electron beam gray scale image, and de-noising and edge extraction iteration can be performed on the superimposed electron beam gray scale image.

[0099] Specifically, the primary electron beam gray scale image profile and the original electron beam gray scale image can be superimposed to obtain a superimposed electron beam gray scale image, that is, the original gray scale image is superimposed after the edge is extracted. The purpose of this step is to strengthen the image edge and make the image more sharp.

[0100] The de-noising and edge extraction iteration is performed, referring to Figure 5 Fig. 3 shows a schematic diagram of an iteration electron beam gray scale image provided by the embodiment of the present application. The image noise can be reduced, the accuracy of edge extraction can be improved, and a more accurate image edge can be obtained.

[0101] S105: when the iteration reaches a preset number of times, performing morphological processing on the iteration electron beam gray scale image obtained by iteration to obtain a target electron beam image edge profile curve image.

[0102] S106: image and operation is performed on the target electron beam image edge profile curve image to obtain a target electron beam single edge image.

[0103] In the embodiment of the present application, when the iteration reaches a preset number of times, the loop is exited and the image is output. The processing method using iteration can improve the accuracy of edge processing, refine the edge, and obtain a more accurate double edge image. After iteration processing, the image range can be accurately obtained, part of the pseudo edge can be removed, and the edge can be limited within the range to facilitate further morphological processing.

[0104] Referring to Figure 6 As shown in the figure, it is a schematic diagram of a target electron beam image edge profile curve image provided by the embodiment of the present application. Morphological processing refers to extracting image components meaningful for expressing and depicting the shape of a region from an image to facilitate further image analysis and target recognition. The target electron beam image edge profile curve image is obtained, that is, the edge extracted from the background-line due to the dramatic change in gray value.

[0105] Finally, the and operation is performed, referring to Figure 7 As shown in the figure, it is a schematic diagram of a target electron beam single edge image provided by the embodiment of the present application. The and operation is performed on each pixel value of two images, and the purpose is to obtain the region of interest from an image.

[0106] In the embodiment, the object of the logical and operation is the edge image after morphological processing and the edge image extracted after only one adaptive filtering processing. The edge of the image after morphological processing is extracted to obtain the target electron beam single edge image. To ensure the accuracy of edge profile extraction, the obtained boundary image is appropriately inflated and the and operation is performed with the initial electron beam gray value profile to obtain the target electron beam single edge image.

[0107] In a possible implementation, the embodiment of the present application provides morphological processing on the iteration electron beam gray value image obtained by iteration to obtain a target electron beam image edge profile curve image, which can specifically include:

[0108] The iteration electron beam gray value image is subjected to a closing operation to obtain an operation electron beam gray value image; the first electron beam gray value image is subjected to hole filling to obtain a filled electron beam gray value image; the connected domains in the filled electron beam gray value image are marked, and the connected domains with an area smaller than a preset value are removed to obtain a target electron beam image edge profile curve image.

[0109] That is, in the embodiment of the present application, the iteration electron beam gray value image is obtained, referring to Figure 5, there are edge breakage, double edge and other problems. When performing SEM image analysis, the background-line intersection is selected as the image edge. Due to the high pixel value near the edge, double edges will be extracted when the edge is extracted. When performing SEM measurement, we need the edge from the background (or, the structure recess) to the line. The edge obtained inside the line needs to be filtered out. The operation of first expansion and then corrosion is used, that is, the closed operation is first performed. The purpose of this step is to fill the small black areas inside the white object, connect adjacent objects, and also smooth the boundary without significantly changing the area. The image is first expanded to expand the bright area and connect the broken edge. Then the image is corroded to remove the false edge and restore the true edge to obtain the electron beam gray scale image. After this operation, the hole filling processing is performed. The purpose of filling the hole is to fill the black area surrounded by the white area inside the line. This step can remove the disconnected domain inside the line to obtain the accurate edge to obtain the filled electron beam gray scale image.

[0110] After morphological processing such as expansion, corrosion, and hole filling, there are still black areas inside the line, which exist at the boundary position of the image. These areas inside the line will affect the edge extraction result, and after edge extraction, false edges will appear, which need to be removed before edge extraction. These black areas are not wrapped by white areas and do not meet the requirements of the hole filling operation. These residual black areas inside the line have the characteristics of short length and small area. In this embodiment, we use the method of marking connected domains to remove the connected domains with small areas in the image, which can remove small areas that do not meet the requirements of the line to obtain the accurate line area range, as shown in Figure 6 , to obtain the target electron beam image edge profile curve image.

[0111] In a possible implementation, the embodiment of the present application provides an original electron beam gray scale image, which can specifically include:

[0112] Acquiring a current electron beam image; determining whether the current electron beam image is a single-channel gray scale image; if yes, taking the current electron beam image as the original electron beam gray scale image; if no, performing gray scale conversion on the current electron beam image to obtain the original electron beam gray scale image.

[0113] Specifically, in the embodiment of the present application, it can be determined whether the current electron beam image is a single-channel gray scale image, and other types of images are converted into gray scale images. In order to facilitate subsequent unified standardization processing.

[0114] In a possible implementation, the embodiment of the present application provides denoising and edge extraction iteration on the superimposed electron beam gray scale image, which can specifically include:

[0115] The superimposed electron beam gray scale image is denoised by using an adaptive median filtering algorithm to obtain a denoised electron beam gray scale image.

[0116] Specifically, the image processing can be performed by selecting an image iterative processing method. Iteration refers to repeatedly performing certain operations on the image and performing a refinement process on the image edge.

[0117] In this embodiment, the superimposed electron beam gray scale image is denoised by using an adaptive median filtering algorithm to obtain a denoised electron beam gray scale image, thereby reducing image noise and improving the accuracy of edge extraction.

[0118] In this embodiment, the superimposed electron beam gray scale image is denoised by using an adaptive median filtering algorithm to obtain a denoised electron beam gray scale image, thereby reducing image noise and improving the accuracy of edge extraction.

[0119] In this embodiment, the relationship between the set iteration number and the loop variable is determined. If the loop variable is less than the iteration preset number, the next iteration processing is performed. When the condition that the loop variable is less than the iteration number is not met, the loop is exited, and the image is output.

[0120] The method comprises the following steps: obtaining an original electron beam gray scale image; calculating the average value and the standard deviation of the gray scale value of the pixel points of the original electron beam gray scale image; determining a filtering interval according to the three-sigma principle by using the average value and the standard deviation, filtering the original electron beam gray scale image according to the filtering interval to obtain a filtered electron beam gray scale image; performing piecewise linear transformation on the filtered electron beam gray scale image to obtain an enhanced electron beam gray scale image; filtering the enhanced electron beam gray scale image by using an adaptive median filtering algorithm to obtain a filtered electron beam gray scale image; performing edge extraction on the filtered electron beam gray scale image by using an edge extraction operator to obtain a primary electron beam gray scale image contour; superimposing the primary electron beam gray scale image contour and the original electron beam gray scale image to obtain a superimposed electron beam gray scale image, and performing denoising and edge extraction iteration on the superimposed electron beam gray scale image; when the iteration reaches a preset number, performing morphological processing on the iteration electron beam gray scale image obtained by iteration to obtain a target electron beam image edge contour curve image; and performing image and operation on the target electron beam image edge contour curve image to obtain a target electron beam single-edge image. Therefore, the method of setting a filtering interval is adopted, the preliminary noise reduction processing is performed according to the three-sigma principle, part of the noise points is removed while the image points are protected, and the efficiency of noise elimination is improved. The partial linear change is performed to enhance the image contrast, and the contour extraction is further made more accurate. The precise single-edge image is obtained by combining the morphological processing and the and operation. The accuracy of the contour extraction of the noisy electron beam image is improved, and the extraction effect is improved.

[0121] Exemplary device

[0122] Referring to Figure 8 As shown in the figure, the device for extracting the contour of the noisy electron beam image provided by the embodiment of the application comprises:

[0123] The acquisition unit 201 is configured to acquire an original electron beam grayscale image, and calculate the average value and the standard deviation of the pixel grayscale value of the original electron beam grayscale image.

[0124] The filtering unit 202 is configured to determine a filtering interval according to the three-sigma principle by using the average value and the standard deviation, filter the original electron beam grayscale image according to the filtering interval, and obtain a filtered electron beam grayscale image.

[0125] The contour unit 203 is configured to perform piecewise linear transformation on the filtered electron beam grayscale image to obtain an enhanced electron beam grayscale image, filter the enhanced electron beam grayscale image by using an adaptive median filtering algorithm to obtain a filtered electron beam grayscale image, and perform edge extraction on the filtered electron beam grayscale image by using an edge extraction operator to obtain a primary electron beam grayscale image contour.

[0126] The superposition unit 204 is configured to superimpose the primary electron beam grayscale image contour and the original electron beam grayscale image to obtain a superimposed electron beam grayscale image, and perform denoising and edge extraction iteration on the superimposed electron beam grayscale image.

[0127] The morphology unit 205 is configured to, when the iteration reaches a preset number of times, perform morphological processing on an iteration electron beam grayscale image obtained by iteration to obtain a target electron beam image edge contour curve image.

[0128] The operation unit 206 is configured to perform image and operation on the target electron beam image edge contour curve image to obtain a target electron beam single-edge image.

[0129] In a possible implementation, the morphology unit is specifically configured to:

[0130] perform a closing operation on the iteration electron beam grayscale image to obtain an operation electron beam grayscale image;

[0131] perform hole filling on the first electron beam grayscale image to obtain a filled electron beam grayscale image;

[0132] mark connected domains in the filled electron beam grayscale image, and remove the connected domains with an area smaller than a preset value to obtain the target electron beam image edge contour curve image.

[0133] In a possible implementation, the acquisition unit is specifically configured to:

[0134] acquire a current electron beam image;

[0135] determine whether the current electron beam image is a single-channel gray image;

[0136] if yes, the current electron beam image is taken as the original electron beam gray image;

[0137] if no, the current electron beam image is subjected to gray conversion to obtain the original electron beam gray image.

[0138] In a possible implementation, the superimposition unit is specifically configured to:

[0139] perform denoising on the superimposed electron beam gray image by using an adaptive median filtering algorithm to obtain a denoised electron beam gray image;

[0140] perform edge extraction on the denoised electron beam gray image by using an edge extraction operator to obtain an edge-extracted electron beam gray image.

[0141] The embodiment of the present application provides a contour extraction device for a noisy electron beam image, and a method applied to the device includes: acquiring an original electron beam gray image; calculating an average value and a standard deviation of pixel gray values of the original electron beam gray image; determining a filtering interval according to a three-sigma principle by using the average value and the standard deviation, filtering the original electron beam gray image according to the filtering interval to obtain a filtered electron beam gray image; performing piecewise linear transformation on the filtered electron beam gray image to obtain an enhanced electron beam gray image; performing filtering on the enhanced electron beam gray image by using an adaptive median filtering algorithm to obtain a filtered electron beam gray image; performing edge extraction on the filtered electron beam gray image by using an edge extraction operator to obtain a primary electron beam gray image contour; superimposing the primary electron beam gray image contour and the original electron beam gray image to obtain a superimposed electron beam gray image, and performing denoising and edge extraction iteration on the superimposed electron beam gray image; when the iteration reaches a preset number of times, performing morphological processing on an iteration electron beam gray image obtained through the iteration to obtain a target electron beam image edge contour curve image; and performing image and operation on the target electron beam image edge contour curve image to obtain a target electron beam single-edge image. Thus, the method of setting a filtering interval is adopted, preliminary noise reduction is performed according to the three-sigma principle, part of the noise points is removed while the image points are protected, and the efficiency of noise elimination is improved. Partial linear change is performed, the image contrast is enhanced, and the contour extraction is further more accurate. The morphological processing is combined, and a precise single-edge image is obtained by using the and operation. The accuracy of the contour extraction of the noisy electron beam image is improved, and the extraction effect is improved.

[0142] On the basis of the above-mentioned embodiments, the present embodiment provides a profile extraction system of a noisy electron beam image, comprising:

[0143] a memory for storing a computer program;

[0144] a processor for implementing the steps of the profile extraction method of the noisy electron beam image as described above when executing the computer program.

[0145] On the basis of the above-mentioned embodiments, the present embodiment further provides a computer readable medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of the profile extraction method of the noisy electron beam image as described above.

[0146] It should be noted that the computer readable medium of the present disclosure as described above can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present disclosure, the computer readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries computer readable program code. Such a propagated data signal can take many forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium, which can send, propagate or transmit a program for use by or in conjunction with an instruction execution system, device or apparatus. The program code contained on the computer readable medium can be transmitted using any suitable medium, including but not limited to a wire, an optical cable, an RF (radio frequency) or the like, or any suitable combination of the above.

[0147] The computer readable medium as described above can be contained in the system as described above; or can exist separately and not be assembled into the system.

[0148] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program in accordance with embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for carrying out the methods illustrated by the flowcharts.

[0149] The various embodiments described in this specification are presented by way of example, and are not intended to limit the scope of the application. The various embodiments described in this specification can be combined or combined in any order to provide other embodiments. Various embodiments described in this specification can be implemented in any combination or sub-combination. The same or similar reference numerals in different drawings represent the same or similar elements.

[0150] The above descriptions are only the preferred embodiments of the present application, and the present application is not limited to the above. Any person skilled in the art can make many possible changes and modifications to the technical solutions of the present application or modify equivalent embodiments without departing from the scope of the technical solutions of the present application. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the scope of the technical solutions of the present application, shall still fall within the scope of protection of the technical solutions of the present application.

Claims

1. A method for contour extraction from a noisy electron beam image, characterized in that, include: Obtain the original electron beam grayscale image; Calculate the mean and standard deviation of the grayscale values ​​of the pixels in the original electron beam grayscale image; The filtering interval is determined according to the three sigma principle using the mean and the standard deviation, and the original electron beam grayscale image is filtered according to the filtering interval to obtain the filtered electron beam grayscale image; The filtered electron beam grayscale image is subjected to piecewise linear transformation to obtain an enhanced electron beam grayscale image; The enhanced electron beam grayscale image is filtered using an adaptive median filtering algorithm to obtain a filtered electron beam grayscale image. Edge extraction is performed on the filtered electron beam grayscale image using an edge extraction operator to obtain the outline of the initial electron beam grayscale image; The initial electron beam grayscale image outline and the original electron beam grayscale image are superimposed to obtain a superimposed electron beam grayscale image. The superimposed electron beam grayscale image is then subjected to denoising and edge extraction iteratively. When the iteration reaches a preset number of times, the iterative electron beam grayscale image obtained by the iteration is subjected to morphological processing to obtain the edge contour curve image of the target electron beam image. The edge contour curve of the target electron beam image is ANDed with the contour of the initial electron beam grayscale image to obtain a single edge image of the target electron beam.

2. The method according to claim 1, characterized in that, The step of performing morphological processing on the iteratively obtained electron beam grayscale image to obtain the target electron beam image edge contour curve image includes: The iterative electron beam grayscale image is closed by a closing operation to obtain the operational electron beam grayscale image; The operational electron beam grayscale image is filled with holes to obtain a filled electron beam grayscale image; The connected components in the filled electron beam grayscale image are marked, and the connected components with an area smaller than a preset value are removed to obtain the edge contour curve image of the target electron beam image.

3. The method according to claim 1, characterized in that, The acquisition of the original electron beam grayscale image includes: Acquire the current electron beam image; Determine whether the current electron beam image is a single-channel grayscale image; If so, the current electron beam image is used as the original electron beam grayscale image; If not, the current electron beam image is converted to grayscale to obtain the original electron beam grayscale image.

4. The method according to claim 1, characterized in that, The iterative denoising and edge extraction of the superimposed electron beam grayscale image includes: The superimposed electron beam grayscale image is denoised using an adaptive median filtering algorithm to obtain a denoised electron beam grayscale image. Edge extraction is performed on the denoised electron beam grayscale image using an edge extraction operator to obtain an edge-extracted electron beam grayscale image.

5. A contour extraction device for noisy electron beam images, characterized in that, include: The acquisition unit is used to acquire the original electron beam grayscale image; Calculate the mean and standard deviation of the grayscale values ​​of the pixels in the original electron beam grayscale image; A filtering unit is used to determine a filtering interval based on the average value and the standard deviation according to the three sigma principle, and to filter the original electron beam grayscale image according to the filtering interval to obtain a filtered electron beam grayscale image. The contour unit is used to perform piecewise linear transformation on the filtered electron beam grayscale image to obtain an enhanced electron beam grayscale image. The enhanced electron beam grayscale image is filtered using an adaptive median filtering algorithm to obtain a filtered electron beam grayscale image. Edge extraction is performed on the filtered electron beam grayscale image using an edge extraction operator to obtain the outline of the initial electron beam grayscale image; The overlay unit is used to overlay the outline of the initial electron beam grayscale image and the original electron beam grayscale image to obtain an overlay electron beam grayscale image, and to perform denoising and edge extraction iteratively on the overlay electron beam grayscale image; A morphological unit is used to perform morphological processing on the iterative electron beam grayscale image obtained by the iteration when the iteration reaches a preset number of times, so as to obtain the edge contour curve image of the target electron beam image. The processing unit is used to perform an AND operation between the edge contour curve image of the target electron beam image and the contour of the initial electron beam grayscale image to obtain a single edge image of the target electron beam.

6. The apparatus according to claim 5, characterized in that, The morphological unit is specifically used for: The iterative electron beam grayscale image is closed by a closing operation to obtain the operational electron beam grayscale image; The operational electron beam grayscale image is filled with holes to obtain a filled electron beam grayscale image; The connected components in the filled electron beam grayscale image are marked, and the connected components with an area smaller than a preset value are removed to obtain the edge contour curve image of the target electron beam image.

7. The apparatus according to claim 5, characterized in that, The acquisition unit is specifically used for: Acquire the current electron beam image; Determine whether the current electron beam image is a single-channel grayscale image; If so, the current electron beam image is used as the original electron beam grayscale image; If not, the current electron beam image is converted to grayscale to obtain the original electron beam grayscale image.

8. The apparatus according to claim 5, characterized in that, The superposition unit is specifically used for: The superimposed electron beam grayscale image is denoised using an adaptive median filtering algorithm to obtain a denoised electron beam grayscale image. Edge extraction is performed on the denoised electron beam grayscale image using an edge extraction operator to obtain an edge-extracted electron beam grayscale image.

9. A contour extraction system for noisy electron beam images, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the contour extraction method for noisy electron beam images as described in any one of claims 1-4.

10. A computer-readable medium, characterized in that, The computer-readable medium stores a computer program that, when processed and executed, implements the steps of the contour extraction method for noisy electron beam images as described in any one of claims 1-4.

Images