A method and system for evaluating the imaging quality of a scanning electron microscope

Abstract

This application discloses an imaging quality evaluation method and system for scanning electron microscopy, relating to the field of image processing technology. The method utilizes a scanning electron microscope to scan the featureless regions of a standard sample that meet set requirements for surface smoothness, compositional uniformity, and electrical stability to obtain electron microscopic images. The electron microscopic images are then divided into blocks to ensure statistical significance of grayscale within each block and comparability of grayscale between blocks. A global grayscale consistency index is calculated based on the spatial distribution differences of grayscale statistical values ​​across the blocks, yielding the imaging quality evaluation result. This method features unified evaluation standards, good repeatability, and objective, accurate, and reliable imaging quality evaluation results. It effectively reflects systematic image distortions that are difficult to detect with traditional resolution, signal-to-noise ratio, and other indicators, exhibiting good repeatability and comparability.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a method and system for evaluating the imaging quality of a scanning electron microscope. Background Technology

[0002] Scanning electron microscopy (SEM) is a core characterization instrument used in materials science, biology, semiconductor industry, and nanotechnology to observe the microscopic morphology and composition of samples. The imaging principle of SEM is as follows: a high-energy electron beam is focused and scans the sample surface, exciting secondary electrons and backscattered electrons. A detector receives these signal electrons and converts them into an electrical signal proportional to the signal intensity, thus obtaining an electron micrograph. Since the number of signal electrons is closely related to the morphological variations and compositional differences of the sample surface, the grayscale value of each point on the electron micrograph can directly and accurately reflect the morphological level or compositional differences of the corresponding point on the sample surface.

[0003] However, in actual imaging processes, the imaging quality of scanning electron microscopes is affected by various factors, including the characteristics of the electron beam itself (such as beam current instability, beam spot distortion, and scanning coil nonlinearity), the state of the detector and electron optical system (such as detector response non-uniformity and signal amplification circuit drift), and sample and environmental factors (such as charging effects, local contamination, and environmental vibration). These interferences introduce different types of distortion into electron micrographs, such as regional brightness non-uniformity, stripe artifacts, geometric distortion, or contrast anomalies. These distortions not only affect the visual effect of electron micrographs but, more seriously, can interfere with or even mislead subsequent quantitative analyses, such as dimensional measurement, particle analysis, and porosity calculation, causing measurement results to deviate from the true values. For applications requiring extremely high precision, such as critical dimension measurement in the semiconductor industry and microstructural characterization in materials science, fluctuations in imaging quality directly affect the reliability of analytical conclusions.

[0004] Therefore, objective and quantitative evaluation of the imaging quality of scanning electron microscopes is of significant practical importance for monitoring the imaging system's status, optimizing image acquisition parameters, and ensuring the reliability of subsequent analysis results. Currently, the main methods for evaluating the imaging quality of scanning electron microscopes are as follows: (1) Resolution evaluation method. This method evaluates the resolving power of a scanning electron microscope by photographing standard samples (such as gold particles, tin balls, etc.) and using Fourier transform, contrast-gradient method or derivative method to assess image sharpness. However, resolution evaluation mainly focuses on the detail resolution limit of the image and emphasizes the high-frequency transmission characteristics of the imaging system. It lacks effective quantitative means for image uniformity problems caused by factors such as electron beam drift, scanning distortion or detector gain non-uniformity.

[0005] (2) Signal-to-noise ratio (SNR) evaluation method. This method involves acquiring two images of the same area and calculating the ratio of the mean gray value to the standard deviation of the selected area in the image, or using methods such as "double image subtraction" to obtain the SNR of the image. This method mainly measures the influence of random noise and is often not sensitive enough to systematic image distortions (such as regional brightness gradients, scanning stripes, etc.), making it difficult to fully reflect the imaging quality.

[0006] (3) Contrast Evaluation Method. This method uses the difference between the maximum and minimum gray levels of feature regions in the image (such as the edges of holes and particle boundaries), or the distribution width of the gray-level histogram, to evaluate the overall contrast of the image. However, this method depends on the presence of obvious structural boundaries in the image, and is less applicable to samples with flat surfaces and sparse features. More importantly, contrast evaluation cannot distinguish whether the observed differences in brightness and darkness originate from the true morphology or composition of the sample itself, or from non-ideal factors introduced by the imaging system, which limits its application in system state assessment.

[0007] (4) Subjective visual evaluation method. This method involves experienced operators directly observing the image and making qualitative judgments based on characteristics such as uniform brightness, presence of shadows or bands, and clarity of details. This method does not require complex calculations and can quickly detect obvious image anomalies. It is the most commonly used and direct method in actual work. However, this method is highly dependent on the operator's subjective experience. The judgments of different personnel, or even the same person at different times, may be inconsistent. The evaluation criteria are difficult to unify and quantify. Moreover, it is often difficult to detect weak and gradual image degradation, and it cannot meet the needs of standardized and automated quality control. Summary of the Invention

[0008] This application addresses the aforementioned problems and technical needs by proposing a method and system for evaluating the imaging quality of a scanning electron microscope. The technical solution of this application is as follows: Firstly, a method for evaluating the imaging quality of a scanning electron microscope, the method comprising: Scanning electron microscopes are used to scan the featureless areas of standard samples to obtain electron microscopic images. The surface flatness, compositional uniformity, and electrical stability of the featureless areas of the standard samples meet the flatness requirements, compositional uniformity requirements, and electrical stability requirements. The electron micrograph is divided into multiple image blocks; Calculate the grayscale statistics of pixels in each image block; The global grayscale consistency index of the electron microscope image is calculated based on the deviation between the grayscale statistical values ​​of each image block, and the imaging quality evaluation result of the scanning electron microscope is obtained. The higher the global grayscale consistency of the electron microscope image, the better the imaging quality of the scanning electron microscope is indicated by the imaging quality evaluation result.

[0009] A further technical solution is that the imaging quality evaluation method for scanning electron microscopy also includes: The lower limit of the number of pixels per block to meet the accuracy requirements for imaging quality evaluation is determined based on the imaging parameters of the scanning electron microscope. , The parameter is an integer. When dividing an electron microscope image into multiple image blocks, the number of pixels in each resulting image block shall not be less than the lower limit of the number of pixels in a single block. .

[0010] A further technical solution is that the imaging parameters of the scanning electron microscope include the electron beam current. Pixel electron beam dwell time Detector electron collection efficiency Based on the imaging parameters of the scanning electron microscope, determine the lower limit of the number of pixels per block to meet the accuracy requirements for imaging quality evaluation. include: Determine the average number of electrons per pixel based on the imaging parameters of a scanning electron microscope. ,in, It is the charge of a single electron; Based on the characteristic that the gray values ​​of individual pixels in electron micrographs follow a Poisson distribution, the inclusion of... The overall relative statistical fluctuation of the total average number of electrons in an image block of pixels is as follows: , The parameter is an integer. exist Determine under constraints ,in, It is a relative statistical fluctuation threshold, and the higher the accuracy requirement for image quality evaluation, the better. The smaller the value, the better.

[0011] A further technical solution involves dividing the electron microscope image into multiple image blocks, including: Based on the lower limit of the number of pixels in a single block The image resolution of an electron microscope image determines the smallest block size of a single image segment. ; Based on the sensitivity requirements and anti-interference capability requirements for imaging quality evaluation, the minimum block size is... and the set maximum block size Within a certain range, the block size is determined, and the electron micrograph is divided into multiple image blocks according to the determined block size; among them, the smaller the block size, the higher the sensitivity of imaging quality evaluation, and the lower the anti-interference ability of imaging quality evaluation.

[0012] A further technical solution involves dividing the electron microscope image into multiple image blocks, including: The electron micrograph is evenly divided into multiple pre-analysis blocks according to the pre-block size, and the gray-scale statistical value of the pixels in each pre-analysis block is calculated. Based on the differences in grayscale statistical values ​​of each pre-analysis block, regions with drastic grayscale changes and regions with gradual grayscale changes in the electron micrograph are identified. Based on the sensitivity requirements and anti-interference capability requirements for imaging quality evaluation, the minimum block size is... and the set maximum block size Within a certain range, the first block size and the second block size are determined respectively. According to the first block size, the area with drastic gray-scale changes in the electron microscope image is divided into multiple image blocks, and according to the second block size, the area with gentle gray-scale changes in the electron microscope image is divided into multiple image blocks. The first block size is smaller than the second block size.

[0013] A further technical solution involves calculating the global grayscale consistency index of the electron microscopy image based on the deviation between the grayscale statistical values ​​of each image block, including: Calculate the global grayscale consistency index of electron micrographs Global grayscale consistency index The smaller the value, the higher the global grayscale consistency of the electron microscope image; in, It is the maximum value among the grayscale statistics of each image block. It is the minimum value among the grayscale statistics of each image block.

[0014] A further technical solution is that the imaging quality evaluation method of this scanning electron microscope also includes: Identify image blocks with abnormal grayscale statistics in electron micrographs to obtain the grayscale aberration regions of electron micrographs. Multiple electron micrographs were obtained by controlling the scanning electron microscope to scan the featureless area of ​​the standard sample multiple times, and the gray-scale abnormal area of ​​each electron micrograph was determined. When the gray-scale abnormal areas of multiple electron micrographs are consistent, it is determined that the imaging quality of scanning electron microscopy is affected by spatial factors. When grayscale anomalies appear in multiple electron micrographs with inter-row fluctuations or random spot-like distribution, it indicates that the imaging quality of the scanning electron microscope is affected by time factors.

[0015] A further technical solution involves the standard sample also including locally characteristic regions, and controlling a scanning electron microscope to scan the non-characteristic regions of the standard sample to obtain electron microscopic images further includes: The scanning electron microscope (SEM) is controlled to scan a localized characteristic area of ​​a standard sample to complete the imaging calibration operation. After completing the imaging calibration operation, the SEM is then controlled to scan a non-characteristic area of ​​the standard sample to obtain an electron micrograph.

[0016] A further technical solution involves dividing the electron microscope image into multiple image blocks, including: After image preprocessing of the electron microscope image, the preprocessed electron microscope image is divided into multiple image blocks. The image preprocessing performed on the electron microscope image includes: removing pixels in the electron microscope image whose gray values ​​exceed a preset gray value range, and / or, using an image denoising algorithm to denoise the electron microscope image.

[0017] In a second aspect, an imaging quality evaluation system for a scanning electron microscope is provided, characterized in that the imaging quality evaluation system for a scanning electron microscope includes a scanning electron microscope, a standard sample, and electronic equipment. The standard sample is placed in the sample chamber of the scanning electron microscope, and the surface flatness, compositional uniformity, and electrical stability of the non-feature area of ​​the standard sample meet the flatness requirements. Electronic devices include a processor and a memory. The memory stores a computer program, which, when executed by the processor, causes the electronic device to perform the method described in the first aspect above.

[0018] The beneficial technical effects of this application are: This application discloses an imaging quality evaluation method and system for scanning electron microscopes. The method utilizes a scanning electron microscope to scan the featureless regions of a standard sample that meet set requirements for surface smoothness, compositional homogeneity, and electrical stability to obtain electron microscopic images. The electron microscopic images are then divided into blocks to ensure statistical significance of grayscale within each block and comparability of grayscale between blocks. The imaging quality of the scanning electron microscope is then objectively and quantitatively evaluated based on the spatial distribution characteristics of the grayscale statistical values ​​of each block. This method effectively reflects systematic image distortions that are difficult to detect using traditional indicators such as resolution and signal-to-noise ratio. The evaluation criteria are unified, and the repeatability is good. The obtained imaging quality evaluation results are objective, accurate, and reliable, with good repeatability and comparability.

[0019] This method also provides a way to quantitatively calculate the lower limit of the number of pixels in a single block. Based on this lower limit, the accurate block size can be determined and the electron microscope image can be divided into blocks, thereby ensuring the accuracy, sensitivity and anti-interference ability of the imaging quality evaluation of scanning electron microscope.

[0020] This method can also locate gray-scale abnormal regions in electron microscopy images based on the gray-scale statistical values ​​of pixels in each image block. Combined with the working characteristics of scanning electron microscopy's line-by-line scanning imaging, based on the distribution of gray-scale abnormal regions in multiple electron microscopy images, it can preliminarily determine whether the interference experienced by the scanning electron microscope is spatial or temporal interference, which can provide a basis for further quality optimization of subsequent interference.

[0021] This method uses an objective and repeatable global grayscale consistency index to evaluate image quality. It requires no complex hardware modifications or expensive software, making it easy to promote and apply. Furthermore, each step can be automated by computer software, significantly improving work efficiency. This method provides a unified evaluation standard for comparing image quality between different laboratories and different scanning electron microscopes, contributing to the establishment and improvement of quality control systems in the scanning electron microscope industry. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating an imaging quality evaluation method for a scanning electron microscope according to one embodiment of this application.

[0023] Figure 2 This is a schematic flowchart of a method for dividing an electron microscope image into multiple image blocks in another embodiment of this application.

[0024] Figure 3 This is a flowchart illustrating a method for determining the lower limit of the number of pixels in a single block, as described in another embodiment of this application.

[0025] Figure 4 This is an example of an electron micrograph obtained by scanning a standard sample 1 using a scanning electron microscope 1 to be evaluated.

[0026] Figure 5 Yes Figure 4 A schematic diagram of an electron micrograph after image segmentation and Viridis color mapping.

[0027] Figure 6 yes Figure 5 A schematic diagram showing the grayscale statistics of each image block and the corresponding Viridis color mapping results.

[0028] Figure 7This is another example of an electron micrograph obtained by scanning a standard sample 2 using a scanning electron microscope 2 to be evaluated.

[0029] Figure 8 Yes Figure 7 A schematic diagram of an electron micrograph after image segmentation and Viridis color mapping.

[0030] Figure 9 yes Figure 8 A schematic diagram showing the grayscale statistics of each image block and the corresponding Viridis color mapping results.

[0031] Figure 10 This is a structural block diagram of an electronic device in an imaging quality evaluation system for a scanning electron microscope according to one embodiment of this application. Detailed Implementation

[0032] The specific embodiments of this application will be further described below with reference to the accompanying drawings.

[0033] This embodiment discloses a method for evaluating the imaging quality of a scanning electron microscope. Please refer to [link / reference]. Figure 1 The flowchart shown illustrates the imaging quality evaluation method for scanning electron microscopy, which includes the following steps: Step 110: Control the scanning electron microscope to scan the featureless area of ​​the standard sample to obtain an electron micrograph.

[0034] The standard sample used in this step is pre-prepared. The surface flatness of the non-feature areas of the standard sample meets the flatness requirements, the uniformity of composition meets the uniformity requirements, and the electrical stability meets the stability requirements, so as to avoid introducing sample noise interference.

[0035] In another embodiment, the surface roughness of the featureless regions of the standard sample does not exceed a roughness threshold and is free of surface defects to meet flatness requirements. Surface defects include scratches, stains, impurities, and oxide layers. The featureless regions of the standard sample are free of compositional segregation and crystal defects within the error range to meet homogeneity requirements. The featureless regions of the standard sample are stable under electron beam irradiation and do not undergo charging effects or irradiation damage within the error range to meet stability requirements. Surface defects include scratches, stains, impurities, and oxide layers. In another embodiment, the roughness threshold is 0.008 μm.

[0036] In another embodiment, the standard sample is made of a single-sided polished monocrystalline silicon wafer or a polished metal sheet, including polished aluminum or polished copper. Before placing the standard sample into the sample chamber of the scanning electron microscope, the standard sample is first cleaned, including ultrasonic cleaning with acetone for 5 minutes, ultrasonic cleaning with anhydrous ethanol for 10 minutes, and finally drying with high-purity nitrogen gas to avoid residual contaminants on the surface affecting the evaluation results.

[0037] In practice, the standard sample is placed on the sample stage in the sample chamber of the scanning electron microscope (SEM). Following the standard operating procedure, the microscope is powered on, vacuumed, and the sample stage is adjusted. Then, the SEM is adjusted to the required operating parameters to scan the featureless areas of the standard sample, obtaining an electron micrograph. The operating parameters of the SEM include accelerating voltage, working distance, and magnification.

[0038] Furthermore, during the imaging process of scanning electron microscopy (SEM), the signal quality generated by the interaction between the high-energy electron beam and the standard sample is directly affected by the focusing state, astigmatism magnitude, and signal acquisition link gain. If the focusing is inaccurate, the electron beam spot cannot converge to its minimum diameter, easily leading to blurred electron micrographs and loss of detail. If uncompensated astigmatism exists, the focusing ability of the electron beam in orthogonal directions is inconsistent, easily causing differential stretching of feature edges in different directions in the electron micrograph, resulting in distortion. Improper brightness / contrast settings may lead to signal saturation clipping or dark details being submerged in a noisy substrate, affecting the complete extraction of effective information. Therefore, before using an SEM to scan the featureless areas of a standard sample to obtain an electron micrograph, necessary imaging calibration operations must be performed on the SEM. This is a necessary prerequisite to ensure the accuracy and reliability of the subsequent electron micrographs.

[0039] The effectiveness of the aforementioned imaging calibration operation depends on the presence of a feedback reference that can be identified by the system in the image. Since the featureless areas of the standard sample lack necessary feedback references, in another embodiment, the standard sample includes featureless areas and locally featured areas. The locally featured areas are regions with identifiable morphological edges or grayscale differences. In actual implementation, the proportion of locally featured areas does not exceed a certain threshold, meaning that the standard sample is mostly uniformly featureless with locally featured areas. After completing operations such as power-on, vacuuming, sample stage adjustment, and working parameter settings, the scanning electron microscope (SEM) is first controlled to scan the locally featured areas of the standard sample to complete the SEM imaging calibration operation. Then, the SEM, having completed the imaging calibration operation, is controlled to scan the featureless areas of the standard sample to obtain an electron micrograph, thereby ensuring that the obtained electron micrograph is clear and has moderate overall brightness. The imaging calibration operations performed on the locally featured areas of the standard sample include focusing, astigmatism correction, and brightness / contrast adjustment. Specific imaging calibration operations can refer to existing methods and will not be elaborated here.

[0040] In another embodiment, to avoid the image sharpness of the electron microscope image affecting the accuracy and precision of subsequent imaging quality evaluation, the image resolution of the electron microscope image acquired in this step is not lower than a preset resolution threshold, and it is stored in a lossless format to avoid lossy compression formats that cause grayscale information loss. In one example, the resolution threshold is 1024×1024 pixels, and a lossless format such as TIF format is used.

[0041] Step 120: Divide the acquired electron micrograph into multiple image blocks.

[0042] The grayscale value of each pixel in an electron micrograph obtained by scanning electron microscopy (SEM) is actually the number of signal electrons collected by the detector during the pixel's residence time. This number is a discrete random variable. Therefore, even if the featureless region of the standard sample meets the required conditions and the high-energy electron beam is absolutely stable, the number of signal electrons collected by each pixel will fluctuate and strictly follow Poisson statistics. This fluctuation is not an error, but an inherent imaging law of SEM. Therefore, for electron micrographs obtained by scanning electron microscopy, it is necessary to first divide the electron micrograph into multiple image blocks, so that the total electron count in each image block reaches a statistically significant level, thereby ensuring the comparability of grayscale values ​​between image blocks. The multiple image blocks cover the entire area of ​​the electron micrograph, and in one embodiment, the multiple image blocks do not overlap.

[0043] In another embodiment, to further eliminate the influence of interference factors, after acquiring the electron microscope image in step 110, image preprocessing is first performed on the acquired electron microscope image, and then the preprocessed electron microscope image is divided into multiple image blocks according to step 120. The image preprocessing performed on the electron microscope image includes: Pixels in the electron microscope image whose grayscale values ​​exceed a preset grayscale value range are removed. This ensures that overly bright pixels (exceeding the preset grayscale value range) and overly dark pixels (below the preset grayscale value range) are not included in subsequent calculations, thus filtering out irrelevant information in the electron microscope image, such as local overexposure or underexposure caused by minute features. The preset grayscale value range is a pre-defined empirical value.

[0044] And / or, image denoising algorithms are used to denoise electron microscope images. The image denoising algorithms used include median filtering, Gaussian filtering, etc., which can remove random noise from electron microscope images.

[0045] Image preprocessing can effectively suppress the effects of image noise and local anomalies, which helps to improve the stability and reliability of subsequent imaging quality evaluation results.

[0046] Step 130: Calculate the grayscale statistics of pixels in each image block.

[0047] The grayscale statistics of each image block are used to characterize the distribution of grayscale values ​​of all pixels within that image block. Various commonly used types of grayscale statistics can be used. In another embodiment, the grayscale statistics of the image block are obtained by calculating at least one of the following: location metric, dispersion metric, distribution morphology metric, and information content metric of the grayscale values ​​of all pixels within the image block. The location metric includes at least one of the mean, median, and mode. The dispersion metric includes at least one of the standard deviation and range. The distribution morphology metric includes at least one of the skewness and kurtosis. The information content metric includes at least one of the information entropy and uniformity. This application does not list all other grayscale statistics, nor does it limit the type of grayscale statistics actually used.

[0048] Step 140: Calculate the global grayscale consistency index of the electron microscope image based on the deviation between the grayscale statistical values ​​of each image block, and obtain the imaging quality evaluation result of the scanning electron microscope.

[0049] The smaller the deviation between the gray-level statistical values ​​of each image block, the higher the global gray-level consistency of the electron microscope image represented by the calculated global gray-level consistency index, and the better the imaging quality evaluation result indicates the imaging quality of the scanning electron microscope.

[0050] In another embodiment, the global grayscale consistency index of the electron micrograph is calculated according to the following formula. :

[0051] in, It is the maximum value among the grayscale statistics of each image block. It is the minimum grayscale statistical value among all image blocks. The global grayscale consistency index is calculated from this. The smaller the value, the higher the global grayscale consistency of the electron micrograph.

[0052] Then, further based on the global grayscale consistency index The values ​​are used to obtain the imaging quality evaluation results of scanning electron microscopy, and the global grayscale consistency index. The smaller the value of , the better the imaging quality of the scanning electron microscope is indicated by the imaging quality evaluation results.

[0053] For example, in one instance, the obtained image quality evaluation results include four levels: excellent, good, acceptable, and unacceptable. These four levels correspond to the global grayscale consistency index. The correspondence between the values ​​of is as follows: when When the imaging quality evaluation result is determined to be excellent, When the imaging quality evaluation result is determined to be good, When the imaging quality evaluation result is determined to be qualified, The imaging quality evaluation result was determined to be unqualified.

[0054] The above global grayscale consistency index The correlation between the value range and the imaging quality evaluation result is merely an example and can be adjusted appropriately according to the specific requirements of different application fields. For example, for applications with high imaging quality requirements (such as nanomaterial characterization and semiconductor device testing), the pass standard can be appropriately increased, that is, the global grayscale consistency index corresponding to pass. The upper limit for its value is smaller. However, for general applications, the acceptance criteria can be appropriately relaxed, that is, the global grayscale consistency index corresponding to acceptance. The upper limit of its value is larger and can be set according to the actual application needs.

[0055] Optionally, to improve the reliability of the imaging quality evaluation results, the scanning electron microscope is controlled to scan the featureless areas of the standard sample multiple times to obtain multiple electron micrographs, and the global grayscale consistency index is obtained respectively. Then, the average value is taken to obtain the final image quality evaluation result to reduce randomness. The number of scans should not be less than 3.

[0056] In summary, this imaging quality evaluation method utilizes a scanning electron microscope to scan the featureless regions of a standard sample to obtain electron microscopic images. These images are then segmented into blocks, and the spatial distribution characteristics of the grayscale statistical values ​​of each block are used to objectively and quantitatively evaluate the imaging quality of the scanning electron microscope. This method effectively reflects systematic image distortions that are difficult to detect using traditional indicators such as resolution and signal-to-noise ratio. Furthermore, the evaluation criteria are standardized and highly repeatable, significantly improving the reliability and efficiency of evaluating scanning electron microscope imaging systems.

[0057] In the above embodiments, step 120 requires dividing the acquired electron microscopy image into multiple image blocks. As described in the embodiments above, this step is mainly to ensure that the total electron count within a single image block reaches a statistically significant level, thereby ensuring the grayscale comparability between image blocks. Therefore, the method of dividing the image blocks has a direct impact on the reliability and accuracy of subsequent scanning electron microscope imaging quality evaluation. One approach is to divide the electron microscopy image into multiple equal-sized, non-overlapping rectangular image blocks based on experience. For example, this can be done by dividing the image block into blocks containing 128*128 pixels or 64*64 pixels, or by dividing it into at least 16 image blocks.

[0058] However, the accuracy of image block division methods set based on experience is often insufficient, which may lead to unreasonable image block division and directly affect the reliability of subsequent image quality evaluation. Therefore, in another embodiment, please refer to... Figure 2 The flowchart shown includes the following steps: Step 210: Determine the lower limit of the number of pixels per block to meet the accuracy requirements for image quality evaluation based on the imaging parameters of the scanning electron microscope. , The parameter is an integer.

[0059] In another embodiment, the imaging parameters of the scanning electron microscope to be used include the electron beam current. Pixel electron beam dwell time Detector electron collection efficiency Determine the lower limit of the number of pixels in a single block. The process includes steps 211 to 214, please refer to the following steps. Figure 3 Flowchart: Step 211: Calculate the average number of electrons per pixel based on the imaging parameters of the scanning electron microscope. :

[0060] in, It is the charge of a single electron.

[0061] Step 212, based on the average number of electrons per pixel It can be determined that it contains The total average number of electrons in an image block of a pixel. , The parameter is an integer.

[0062] Step 213: Based on the imaging principle of scanning electron microscopy, it can be determined that the gray value of a single pixel in an electron micrograph strictly follows a Poisson distribution. The characteristic of the Poisson distribution is that the average number of electrons per pixel is... variance of the number of electrons per pixel satisfy This allows us to determine the relative statistical fluctuations in pixel grayscale values. .

[0063] Similarly, for the entire image block, based on the characteristic that the gray values ​​of individual pixels in an electron microscope image follow a Poisson distribution, the total average number of electrons in the block can also be used as a basis. The overall relative statistical fluctuation of the total average number of electrons in the image block is determined as follows: .

[0064] Step 214: Determine the relative statistical fluctuation threshold based on the imaging quality evaluation accuracy requirements. Afterwards, Under the given constraints, the number of pixels within an image block can be determined. The range of values ​​is That is, determining the lower limit of the number of pixels in a single block. .

[0065] Among them, the higher the accuracy requirement for image quality evaluation, the higher the relative statistical fluctuation threshold. The smaller the value, the better. The specific value can be set according to the accuracy requirements of image quality evaluation, such as taking the relative statistical fluctuation threshold. For example, 1%, 2%, etc.

[0066] Step 220, then based on the determined lower limit of the number of pixels per block. To determine the block size, ensure that when dividing an electron microscope image into multiple image blocks, each resulting image block contains no fewer than the lower limit of the number of pixels per block. This is to ensure that the total electron count within a single image block reaches a statistically significant level.

[0067] The requirement is that the number of pixels in each image block is not less than the lower limit of the number of pixels in a single block. To ensure grayscale comparability between image blocks, the block size also affects the sensitivity and anti-interference capability of the method: smaller block sizes result in higher sensitivity to grayscale inhomogeneity and local grayscale differences, but also greater susceptibility to Poisson noise. Conversely, larger block sizes lead to better statistical stability and stronger anti-interference capability, but reduce sensitivity to subtle grayscale inhomogeneity. Sensitivity here refers to the minimum detectable grayscale inhomogeneity, while anti-interference capability refers to the repeatability of grayscale statistics within an image block, i.e., the variance of grayscale statistics within the same image block obtained from multiple imaging and segmentation under the same conditions.

[0068] In another embodiment, step 220 above is based on a lower limit value for the number of pixels in a single block. Determining the block size involves the following steps, please refer to [the relevant documentation]. Figure 2 : Step 221, based on the lower limit of the number of pixels in a single block. The image resolution of the electron microscope image determines the smallest block size of a single image segment. .

[0069] Step 222: Based on the imaging quality evaluation sensitivity requirements and anti-interference capability requirements, within the minimum block size... and the set maximum block size Determine the block size within the specified range.

[0070] Step 223: Divide the electron micrograph into multiple image blocks according to the determined block size.

[0071] Each resulting image block can be either regular or irregular in shape. Smaller block sizes result in higher image quality assessment sensitivity but lower anti-interference capability. In other words, higher sensitivity requirements necessitate larger block sizes, which inevitably sacrifice some anti-interference capability. Conversely, higher anti-interference requirements require larger block sizes, which again sacrifice some sensitivity. The maximum block size is set... The maximum block size can be determined based on the minimum sensitivity requirement for image quality assessment, or empirically by taking the smallest spatial scale of expected non-uniformity in the image, such as taking 1 / 4 of the image side length of an electron micrograph as the maximum block size. The image side length.

[0072] When dividing an electron microscope image into multiple image blocks according to a defined block size, one approach is to determine a single block size, resulting in all image blocks having the same block size. However, in another embodiment, multiple block sizes can be determined to employ a non-uniform division method, including: First, the electron micrograph is evenly divided into multiple pre-analysis blocks according to a pre-block size, and the grayscale statistics of the pixels in each pre-analysis block are calculated. The pre-block size here does not exceed a size threshold, that is, the pre-block size is a pre-set small size, such as a pre-block size containing 16*16 pixels.

[0073] Then, based on the differences in grayscale statistical values ​​of each pre-analysis block, regions with drastic grayscale changes and regions with gradual grayscale changes in the electron micrograph are determined. One approach is to calculate the local variance or local gradient between the grayscale statistical value of each pre-analysis block and the grayscale statistical value of adjacent pre-analysis blocks. Regions where the local variance or local gradient reaches a set threshold are classified as regions with drastic grayscale changes, while regions where the local variance or local gradient does not reach the set threshold are classified as regions with gradual grayscale changes.

[0074] Finally, based on the image quality evaluation sensitivity requirements and anti-interference capability requirements, the minimum block size was determined. and the set maximum block size Within a defined range, a first block size and a second block size are determined. Regions with drastic grayscale changes in the electron microscope image are divided into multiple image blocks according to the first block size, while regions with gradual grayscale changes are divided into multiple image blocks according to the second block size. The first block size is smaller than the second block size. This method results in smaller image blocks in regions with drastic grayscale changes to preserve detail and avoid smoothing out true inhomogeneities with large blocks. Conversely, larger image blocks are used in regions with gradual grayscale changes to improve statistical stability, resulting in a better overall performance in terms of image quality evaluation sensitivity and anti-interference capability.

[0075] In another embodiment, besides calculating the global grayscale consistency index of the electron microscope image according to step 140 above, the deviation between the grayscale statistical values ​​of each image block can also be used to horizontally compare the grayscale statistical values ​​of each image block to identify image blocks with abnormal grayscale statistical values ​​as grayscale aberration regions in the electron microscope image. These grayscale aberration regions of the electron microscope image are the areas with abnormal imaging quality. As mentioned above, the imaging quality of a scanning electron microscope is affected by various factors.

[0076] Unlike the area array imaging of ordinary cameras, scanning electron microscopes (SEMs) employ a line-by-line scanning imaging method. Therefore, SEMs can be considered as using two-dimensional imaging that combines spatial and temporal domains. Based on this imaging characteristic, the SEM is controlled to scan the featureless regions of a standard sample multiple times to obtain multiple electron micrographs. The gray-scale anomaly regions of each electron micrograph are then determined according to the method provided in the above embodiments. By comparing the gray-scale anomaly regions of multiple electron micrographs, when the gray-scale anomaly regions of multiple electron micrographs are consistent, that is, when multiple electron micrographs always have gray-scale anomaly regions in certain fixed locations (such as edges and corners), it is determined that the imaging quality of the SEM is affected by spatial factors. Spatial factors include abnormalities in the electron optical system, abnormal detector position, and abnormal sample stage. When the gray-scale anomaly regions of multiple electron micrographs show inter-row fluctuations or random spot-like distribution, it is determined that the imaging quality of the SEM is affected by temporal factors. Temporal factors include scanning coil nonlinearity, detector response non-uniformity, electron beam instability, and environmental vibration.

[0077] When it is determined that the imaging quality of the scanning electron microscope is low due to interference from external factors, this embodiment can further utilize the different manifestations of various interferences in electron micrographs to investigate interference. In addition to obtaining the imaging quality evaluation results, it can also obtain the preliminary interference investigation results.

[0078] In one example, a single-sided polished monocrystalline silicon wafer was selected as standard sample 1. The surface flatness, compositional uniformity, and electrical stability of the featureless areas of standard sample 1 all meet the testing requirements of this application. Then, standard sample 1 was placed in the sample chamber of scanning electron microscope 1 and evacuated to... The acceleration voltage is set to 10kV, the working distance to 3mm, and the magnification to 3000.

[0079] After performing focusing, astigmatism correction, and other imaging calibration operations on a localized characteristic area of ​​standard sample 1 using scanning electron microscope 1, a featureless area of ​​standard sample 1 was scanned to obtain an electron micrograph with an image resolution of 1536×1536 pixels in TIF format, as shown below. Figure 4 As shown.

[0080] First, the electron micrograph obtained by scanning electron microscope 1 is preprocessed to remove pixels with gray values ​​exceeding the preset gray value range [50, 200]. Then, the block size is determined according to the method provided in this application, and the preprocessed electron micrograph is divided into 25 closely arranged, equal-sized, non-overlapping image blocks of 5*5 pixels according to the block size. To facilitate intuitive observation of the results, this example will... Figure 5The color mapping of the electron micrograph is viridis, and the resulting image block structure is as follows: Figure 5 As shown, based on Figure 5 As can be seen from the viridis color mapping color stripe, Figure 5 In the middle, the closer the color of a pixel is to yellow, the closer its grayscale value is to 255; the closer the color of a pixel is to dark purple, the closer its grayscale value is to 0.

[0081] Then, the grayscale statistics for each image block are calculated. In this example, the average grayscale value of each pixel within each image block is used as the grayscale statistics. The calculation results are as follows: Figure 6 As shown, the number within each image block is its grayscale statistical value, and the color of the image block is the viridis color corresponding to the grayscale statistical value of that image block.

[0082] Therefore, it can be determined that the maximum value among the grayscale statistics of the 25 image blocks is the grayscale statistics of the image block at position A4. The minimum grayscale value among the 25 image blocks is the grayscale value of the image block at position E3. Further calculation yields the global grayscale consistency index D = (129.21 - 126.99) / 126.99 × 100% = 1.75%. Based on the rating method exemplified above, the corresponding imaging quality evaluation result indicates that the imaging quality of the scanning electron microscope 1 is excellent.

[0083] In another example, a single-sided polished monocrystalline silicon wafer was selected as standard sample 2. The surface flatness, compositional uniformity, and electrical stability of the featureless areas of standard sample 2 all met the testing requirements of this application. Then, standard sample 2 was placed in the sample chamber of scanning electron microscope 2 and evacuated to... The following settings are used: accelerating voltage 10kV, reverse field voltage 8kV, working distance 3mm, and amplification factor 4000.

[0084] After performing focusing, astigmatism correction, and other imaging calibration operations on the characteristic areas of standard sample 2 using scanning electron microscope 2, the featureless areas of standard sample 2 were scanned to obtain electron micrographs with an image resolution of 1536×1536 pixels in TIF format, such as... Figure 7 As shown.

[0085] First, the electron micrographs obtained by scanning electron microscope 2 are preprocessed to remove pixels with gray values ​​exceeding the preset gray value range [100, 200], thereby eliminating the influence of fine features on the surface of standard sample 2. Then, the block size is determined according to the method provided in this application, and the preprocessed electron micrographs are divided into 64 closely arranged, equal-sized, non-overlapping image blocks of 8*8 pixels according to the block size. For easier and more intuitive observation of the results, this example will... Figure 7 The color mapping of the electron micrograph was changed to Viridis, and the resulting image block structure is as follows. Figure 8 As shown, Figure 8 The meaning of the viridis color mapping band in the text is the same as above. Figure 5 This will not be elaborated upon here.

[0086] Then, the grayscale statistics for each image block are calculated. In this example, the average grayscale value of each pixel within each image block is used as the grayscale statistics. The calculation results are as follows: Figure 9 As shown, the number within each image block is its grayscale statistical value, and the color of the image block is the viridis color corresponding to the grayscale statistical value of that image block.

[0087] Therefore, it can be determined that the maximum value among the grayscale statistics of the 64 image blocks is the grayscale statistics of the image block at position H8. The minimum grayscale value among the 64 image blocks is the grayscale value of the image block at position G2. Further calculation yields the global grayscale consistency index D = (187.44 – 136.40) / 136.40 × 100% = 37.4%. Based on the rating method exemplified above, the corresponding imaging quality evaluation result indicates that the imaging quality of the scanning electron microscope 2 is unqualified.

[0088] This application also discloses an imaging quality evaluation system for a scanning electron microscope (SEM), which includes an SEM to be evaluated, a standard sample, and electronic equipment. The standard sample is placed in the sample chamber of the SEM, and the surface flatness of the non-feature areas of the standard sample meets the flatness requirement, the compositional homogeneity meets the homogeneity requirement, and the electrical stability meets the stability requirement. Other characteristics of the standard sample can be referred to the descriptions in the above embodiments, and will not be repeated here.

[0089] Please refer to Figure 10The electronic device 1000 includes a processor 1010 and a memory 1020, which communicate with each other via a bus. The memory 1020 stores a computer program, which, when executed by the processor 1010, causes the electronic device 1000 to perform the imaging quality evaluation method in the above embodiments to evaluate the imaging quality of the scanning electron microscope in the imaging quality evaluation system.

[0090] Based on the foregoing description in conjunction with the accompanying drawings, those skilled in the art will understand that the embodiments of this application can also be implemented by software programs. Therefore, this application also provides a computer-readable storage medium. This computer-readable storage medium stores a computer program that, when executed by one or more processors, implements the imaging quality evaluation methods of the above embodiments.

[0091] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the prior art, can be embodied in the form of software products. These computer software products can be stored in computer-readable storage media, such as ROM / RAM, magnetic disks, optical disks, etc., and include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or certain parts of embodiments.

[0092] The above are merely preferred embodiments of this application, and this application is not limited to the above embodiments. It is understood that other improvements and variations that can be directly derived or conceived by those skilled in the art without departing from the spirit and concept of this application should be considered to be included within the protection scope of this application.

Claims

1. A method for evaluating the imaging quality of a scanning electron microscope, characterized in that, The imaging quality evaluation method for the scanning electron microscope includes: Scanning electron microscopes are used to scan the featureless areas of a standard sample to obtain electron microscopic images. The surface flatness, compositional uniformity, and electrical stability of the featureless areas of the standard sample meet the flatness requirements, the compositional uniformity requirements, and the electrical stability requirements. The electron micrograph is divided into multiple image blocks; Calculate the grayscale statistics of pixels in each image block; The global grayscale consistency index of the electron microscope image is calculated based on the deviation between the grayscale statistical values ​​of each image block, and the imaging quality evaluation result of the scanning electron microscope is obtained. The higher the global grayscale consistency of the electron microscope image, the better the imaging quality of the scanning electron microscope is indicated by the obtained imaging quality evaluation result.

2. The method for evaluating the imaging quality of a scanning electron microscope according to claim 1, characterized in that, The imaging quality evaluation method for the scanning electron microscope also includes: Based on the imaging parameters of the scanning electron microscope, determine the lower limit of the number of pixels per block to meet the accuracy requirements for imaging quality evaluation. , The parameter is an integer. When dividing the electron microscope image into multiple image blocks, the number of pixels in each resulting image block is not less than the lower limit of the number of pixels in a single block. .

3. The method for evaluating the imaging quality of a scanning electron microscope according to claim 2, characterized in that, The imaging parameters of the scanning electron microscope include the electron beam current. Pixel electron beam dwell time Detector electron collection efficiency The step involves determining the lower limit of the number of pixels in a single block to meet the accuracy requirements for image quality evaluation based on the imaging parameters of the scanning electron microscope. include: Determine the average number of electrons per pixel based on the imaging parameters of a scanning electron microscope. ,in, It is the charge of a single electron; Based on the characteristic that the gray values ​​of individual pixels in electron micrographs follow a Poisson distribution, the inclusion of... The overall relative statistical fluctuation of the total average number of electrons in an image block of pixels is as follows: , The parameter is an integer. exist Determine under constraints ,in, It is a relative statistical fluctuation threshold, and the higher the accuracy requirement for image quality evaluation, the better. The smaller the value, the better.

4. The method for evaluating the imaging quality of a scanning electron microscope according to claim 2, characterized in that, Dividing the electron micrograph into multiple image blocks includes: Based on the lower limit of the number of pixels in a single block The smallest block size of a single image segment is determined by the image resolution of the electron microscope image. ; Based on the imaging quality evaluation sensitivity requirements and anti-interference capability requirements, in the minimum block size and the set maximum block size Within a certain range, the block size is determined, and the electron micrograph is divided into multiple image blocks according to the determined block size; wherein, the smaller the block size, the higher the sensitivity of the imaging quality evaluation, and the lower the anti-interference ability of the imaging quality evaluation.

5. The method for evaluating the imaging quality of a scanning electron microscope according to claim 4, characterized in that, Dividing the electron micrograph into multiple image blocks also includes: The electron micrograph is evenly divided into multiple pre-analysis blocks according to the pre-block size, and the grayscale statistical value of the pixels in each pre-analysis block is calculated. Based on the differences in grayscale statistical values ​​of each pre-analysis block, the regions with drastic grayscale changes and the regions with gradual grayscale changes in the electron micrograph are determined. Based on the imaging quality evaluation sensitivity requirements and anti-interference capability requirements, in the minimum block size and the set maximum block size Within a certain range, a first block size and a second block size are determined respectively. The region with drastic grayscale changes in the electron microscope image is divided into multiple image blocks according to the first block size, and the region with gentle grayscale changes in the electron microscope image is divided into multiple image blocks according to the second block size. The first block size is smaller than the second block size.

6. The method for evaluating the imaging quality of a scanning electron microscope according to claim 1, characterized in that, The global grayscale consistency index of the electron micrograph is calculated based on the deviation between the grayscale statistical values ​​of each image block, including: Calculate the global grayscale consistency index of the electron micrograph. Global grayscale consistency index The smaller the value, the higher the global grayscale consistency of the electron micrograph; in, It is the maximum value among the grayscale statistics of each image block. It is the minimum value among the grayscale statistics of each image block.

7. The method for evaluating the imaging quality of a scanning electron microscope according to claim 1, characterized in that, The imaging quality evaluation method for the scanning electron microscope also includes: Identify image blocks with abnormal grayscale statistical values ​​in the electron microscope image to obtain the grayscale abnormal region of the electron microscope image; Multiple electron micrographs are obtained by controlling a scanning electron microscope to scan the featureless area of ​​the standard sample multiple times, and the gray-scale abnormal area of ​​each electron micrograph is determined. When the gray-scale abnormal areas of multiple electron micrographs are consistent, it is determined that the imaging quality of the scanning electron microscope is affected by spatial factors. When grayscale anomalies appear in multiple electron micrographs with inter-row fluctuations or random spot-like distribution, it is determined that the imaging quality of the scanning electron microscope is affected by time factors.

8. The method for evaluating the imaging quality of a scanning electron microscope according to claim 1, characterized in that, The standard sample also includes locally characteristic regions, and the process of controlling the scanning electron microscope to scan the non-characteristic regions of the standard sample to obtain electron microscopic images further includes: The scanning electron microscope is controlled to scan a localized characteristic area of ​​the standard sample to complete the imaging calibration operation of the scanning electron microscope. After completing the imaging calibration operation, the scanning electron microscope is controlled to scan a non-characteristic area of ​​the standard sample to obtain an electron micrograph.

9. The method for evaluating the imaging quality of a scanning electron microscope according to claim 1, characterized in that, Dividing the electron micrograph into multiple image blocks includes: After image preprocessing of the electron microscope image, the preprocessed electron microscope image is divided into multiple image blocks; wherein, the image preprocessing performed on the electron microscope image includes: removing pixels in the electron microscope image whose gray values ​​exceed a preset gray value range, and / or, performing noise reduction processing on the electron microscope image using an image denoising algorithm.

10. An imaging quality evaluation system for a scanning electron microscope, characterized in that, The imaging quality evaluation system for the scanning electron microscope includes a scanning electron microscope, standard samples, and electronic equipment; The standard sample is placed in the sample chamber of the scanning electron microscope, and the surface flatness, compositional uniformity, and electrical stability of the non-feature areas of the standard sample meet the flatness requirements, the compositional uniformity requirements, and the electrical stability requirements. The electronic device includes a processor and a memory, the memory storing a computer program that, when executed by the processor, causes the electronic device to perform the method as described in any one of claims 1-9.

Images