Method and system for detecting and identifying real defects by enhanced electron beams

Abstract

The invention relates to a method and system for detecting and identifying real defects by enhancing electron beams, and the method comprises the steps: carrying out the preprocessing of an input electron beam detection image, and extracting a target layer containing a suspected defect region; providing a reference image matched with the target layer, and respectively endowing each pixel with a gray-scale value based on a suspected defect area of the target layer and a brightness distribution condition of a corresponding non-defect area in the reference image; calculating a first absolute value of a gray scale difference value between each pixel in the suspected defect area and a corresponding pixel in the non-defect area, and calculating a first average value based on the first absolute values of all the pixels in the suspected defect area; for different defect categories, collecting a plurality of manually labeled detection samples, and determining a discrimination threshold between different defect categories based on the detection samples; and outputting a defect discrimination result based on a magnitude relationship between the first average value and a discrimination threshold value. According to the invention, the discrimination precision of real defects in the electron beam detection image is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of integrated circuit defect detection, and in particular to a method and system for enhancing electron beam detection to identify real defects. BACKGROUND

[0002] With the rapid development of the semiconductor manufacturing and integrated circuit industry, the process requirements of semiconductor devices are becoming higher and higher. In view of the connection hole communication condition of the product, the wafer foundry (FAB) usually uses an electron beam detection (EBI) machine to monitor the surface defects.

[0003] Electron beam defect detection mainly compares the light and dark differences of the same parts in the detection image and the reference image to distinguish real defects and artifacts. If the defect area is much darker than the surrounding area, it is easy to distinguish, but there are also cases where the surrounding light and dark are very close. When the gray value difference between the defect area and the background is too small, the human eye is often difficult to distinguish, which may cause misjudgment, and the product may fail, thereby affecting the yield. The consistency of cross-product detection may also be different, which becomes one of the key bottlenecks restricting the yield climbing of advanced processes. SUMMARY

[0004] The purpose of the present application is to provide a method and system for enhancing electron beam detection to identify real defects, to solve the problem that the existing electron beam defect detection will cause misjudgment when the gray value difference between the defect area and the background is too small.

[0005] In order to achieve the above purpose, the present application provides a method for enhancing electron beam detection to identify real defects, comprising:

[0006] Pretreating the input electron beam detection image to extract a target layer containing a suspected defect area;

[0007] Providing a reference image matching the target layer, and assigning a gray scale value to each pixel based on the brightness distribution of the suspected defect area of the target layer and the corresponding defect-free area of the reference image;

[0008] Calculating the first absolute value of the gray scale difference between each pixel of the suspected defect area and the corresponding pixel of the defect-free area, and calculating a first average value based on the first absolute values of all pixels of the suspected defect area;

[0009] Collecting a number of artificially labeled detection samples for different defect categories, and determining the discrimination threshold between different defect categories based on the detection samples;

[0010] Outputting the defect discrimination result of the suspected defect area based on the size relationship between the first average value and the discrimination threshold.

[0011] Optionally, the preprocessing comprises:

[0012] peeling off background information of the electron beam detection image, and reserving a circuit pattern region with a circuit structure;

[0013] performing layer separation on the circuit pattern region, and reserving a target layer containing the suspected defect region.

[0014] Optionally, determining the discrimination threshold value between different defect categories based on the detection samples comprises:

[0015] calculating, for each labeled defect region in the detection samples, a second absolute value of a gray scale difference value between each pixel in the defect region and a corresponding pixel in a reference image, and calculating a second average value and a standard deviation based on the second absolute values of all pixels in the defect region;

[0016] performing statistical distribution analysis on the second average value and the standard deviation of the detection samples of different defect categories to obtain a probability density curve of the gray scale difference value distribution of each category;

[0017] taking an intersection point of the probability density curves of the gray scale difference value distribution of different categories as the discrimination threshold value.

[0018] Optionally, the defect categories include dark voltage contrast defects, weak voltage contrast defects, and artifacts, and the discrimination threshold value includes a first threshold value and a second threshold value.

[0019] outputting a defect discrimination result of the target layer based on a size relationship between the first absolute value and the discrimination threshold value comprises:

[0020] when the first absolute value is less than the first threshold value, determining that the artifact;

[0021] when the first absolute value is greater than or equal to the first threshold value and less than the second threshold value, determining that the weak voltage contrast defect;

[0022] when the first absolute value is greater than or equal to the second threshold value, determining that the dark voltage contrast defect.

[0023] Optionally, the detection samples are target layers of different product types and of the same product type under different detection conditions.

[0024] Based on the same inventive concept, the application further provides a system for enhancing electron beam detection and identification of real defects, comprising:

[0025] a extraction module configured to preprocess an input electron beam detection image, and extract a target layer containing a suspected defect region;

[0026] an assignment module configured to assign a gray scale value to each pixel in the suspected defect region and the corresponding non-defect region of the reference image based on a brightness distribution of the suspected defect region and the corresponding non-defect region of the reference image;

[0027] a calculation module configured to calculate a first absolute value of a gray scale difference between each pixel in the suspected defect region and the corresponding pixel in the non-defect region, and calculate a first average value based on the first absolute values of all pixels in the suspected defect region;

[0028] a threshold acquisition module configured to collect a plurality of artificially labeled detection samples for different defect categories, and determine a discrimination threshold between different defect categories based on the detection samples;

[0029] a discrimination module configured to output a defect discrimination result of the suspected defect region based on a size relationship between the first average value and the discrimination threshold.

[0030] Optionally, the pre-processing of the input electron beam detection image by the extraction module includes:

[0031] peeling off background information of the electron beam detection image to retain a circuit pattern region having a circuit structure;

[0032] performing layer separation on the circuit pattern region to retain a target layer containing the suspected defect region.

[0033] Optionally, the determination of the discrimination threshold between different defect categories by the threshold acquisition module based on the detection samples includes:

[0034] calculating a second absolute value of a gray scale difference between each pixel in a labeled defect region of each detection sample and a corresponding pixel of a reference image, and calculating a second average value and a standard deviation based on the second absolute values of all pixels in the defect region;

[0035] performing statistical distribution analysis based on the second average values and the standard deviations of the detection samples of different defect categories to obtain a probability density curve of gray scale difference distribution of each category;

[0036] taking an intersection point of the probability density curves of gray scale difference distribution of different categories as the discrimination threshold.

[0037] Optionally, the defect categories include dark voltage contrast defects, weak voltage contrast defects, and artifacts, and the discrimination threshold includes a first threshold and a second threshold;

[0038] the output of the defect discrimination result of the target layer by the discrimination module based on the size relationship between the first absolute value and the discrimination threshold includes:

[0039] when the first absolute value is greater than or equal to the second threshold value, the dark voltage contrast defect is determined.

[0040] when the first absolute value is greater than or equal to the second threshold value, the dark voltage contrast defect is determined.

[0041] when the first absolute value is greater than or equal to the second threshold value, the dark voltage contrast defect is determined.

[0042] Based on the same inventive concept, the present application also provides a readable storage medium having a computer program stored thereon, the computer program being executable to implement the method for enhancing electron beam detection to identify real defects as described above.

[0043] In the method and system for enhancing electron beam detection to identify real defects provided by the present application, the target layer containing the suspected defect area is peeled off in advance, then the gray scale of the suspected defect area and the corresponding defect-free area in the reference image is assigned according to the brightness distribution, and the average value of the gray scale difference of all pixels in the suspected defect area is calculated, and finally the average value is discriminated by using the discrimination threshold value between different defect categories determined in advance, which effectively improves the discrimination accuracy of real defects in the electron beam detection image, and is suitable for defect detection under different product models / different detection conditions, improves the detection efficiency, and further improves the yield of products. BRIEF DESCRIPTION OF DRAWINGS

[0044] Those skilled in the art should understand that the provided drawings are used to better understand the present application, and do not constitute any limitation on the scope of the present application. Among them:

[0045] Figure 1 a schematic diagram of a dark voltage contrast defect (DVC) provided by an embodiment of the present application;

[0046] Figure 2 a schematic diagram of a weak voltage contrast defect (WVC) provided by an embodiment of the present application;

[0047] Figure 3 a flowchart of the method for enhancing electron beam detection to identify real defects provided by an embodiment of the present application;

[0048] Figure 4 a schematic diagram of the layer peeling result provided by an embodiment of the present application.

[0049] Among them:

[0050] 1-defect area; 2-background information; 3-circuit pattern area; 4-suspected defect area. DETAILED DESCRIPTION

[0051] On the wafer, a large number of Die (chip units) are repeatedly arranged in physical space, and the structure and layout are highly similar. The functional area, circuit pattern level, CT hole and other structures of each Die are almost one-to-one corresponding. Therefore, the same structure should be consistent in different Die. In the electron beam defect detection process, the detection machine usually outputs two images: one is the detection image, including the area suspected of having defects; the other is the reference image, used as a reference benchmark to eliminate background interference and equipment deviation. The comparison image is the image of the side die (chip) of the detection image. If Die A and Die B are left and right adjacent, both have the same pattern structure, if the image of Die A is dark (dark spot) at this position, and Die B is bright and clear, it can be inferred that Die A has defects.

[0052] Both images are black and white grayscale images, and the grayscale value of each pixel ranges from 0 (black) to 255 (white), which physically reflects the strength of the reflection after the electron beam penetrates the structure. As shown in Figure 1 If the defect area 1 (the part selected by the red circle) is much darker than the surrounding area, it is easy to distinguish in this case, but there may also be a situation where the surrounding light and dark are very close, as shown in Figure 2 When the grayscale value difference between the defect area 1 and the background is too small, the human eye is often difficult to distinguish, which may cause misjudgment, especially in CT hole structures. If the hole is not punched, it will appear as a local weak dark spot in the image, usually only 2-5 pixels, the defect volume is extremely small, and the identification difficulty is high. Moreover, the situation may also be different between different products, resulting in poor consistency of cross-product detection, which has become one of the key bottlenecks restricting the yield climbing of advanced processes.

[0053] Therefore, the present application aims to provide a method and system for enhancing the identification of real defects by electron beam detection, which improves the accuracy of real defect identification by using layer peeling, gray scale assignment and threshold discrimination, and is suitable for defect detection under different product models / different detection conditions, improves the detection efficiency, and further improves the yield of the product.

[0054] In order to make the objects, advantages and features of the present application more clearly, the following further describes the present application in conjunction with the drawings and specific embodiments. It should be noted that the drawings are very simplified and all use non-precise proportions, only for the purpose of facilitating, clearly assisting the description of the embodiments of the present application. In order to make the objects, features and advantages of the present application more obvious and easy to understand, please refer to the drawings. It should be noted that the structure, proportion, size, etc. shown in the drawings attached to the present specification are only used to cooperate with the content disclosed in the specification, for those skilled in the art to understand and read, and are not used to limit the defined conditions for implementing the present application. Any modification of structure, change of proportion relationship or adjustment of size, as long as it is the same or similar to the effect and purpose that can be achieved by the present application, should still fall within the scope of the technical content disclosed by the present application.

[0055] As used in the present application, the singular forms "a", "an" and "the" include plural referents unless the content clearly dictates otherwise. As used in the present application, the term "or" is generally used in the sense of "and / or" unless the content clearly dictates otherwise. As used in the present application, the term "several" is generally used in the sense of "at least one" unless the content clearly dictates otherwise. As used in the present application, the term "at least two" is generally used in the sense of "two or more" unless the content clearly dictates otherwise. In addition, the terms "first", "second", "third" are only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second", "third" can explicitly or implicitly include one or at least two features.

[0056] Referring to Figure 3 The present embodiment provides a method for enhancing electron beam detection and identification of real defects, comprising the following steps:

[0057] S1, pre-processing the input electron beam detection image to extract a target layer containing suspected defect area;

[0058] S2, providing a reference image matched with the target layer, and assigning a gray scale value to each pixel based on the brightness distribution of the suspected defect area of the target layer and the corresponding defect-free area in the reference image;

[0059] S3, calculating the first absolute value of the gray scale difference between each pixel of the suspected defect area and the corresponding pixel in the defect-free area, and calculating the first average value based on the first absolute values of all pixels in the suspected defect area;

[0060] S4, for different defect categories, collect several manually annotated detection samples, determine the discrimination threshold between different defect categories based on the detection samples;

[0061] S5, output the defect discrimination result of the suspected defect region based on the size relationship between the first average value and the discrimination threshold.

[0062] Firstly, S1 is performed, the input electron beam detection image is preprocessed, and the target layer containing the suspected defect region is extracted. Further, as shown in Figure 4 The preprocessing includes:

[0063] The background information 2 of the electron beam detection image is stripped, and the circuit pattern area 3 with circuit structure is retained;

[0064] The circuit pattern area 3 is separated by layer, and the target layer containing the suspected defect region 4 is retained.

[0065] In this embodiment, image processing means such as edge detection and morphological operation can be used to strip the invalid background information 2 and retain the area with actual circuit structure to obtain the circuit pattern area 3. Then the circuit pattern area 3 is separated by using structure comparison and hierarchical features to retain the target layer containing the suspected defect region 4 to avoid interference caused by overlapping of different circuit pattern layers.

[0066] Then S2 is performed, a reference image matched with the target layer is provided, and a gray scale value is assigned to each pixel based on the brightness distribution of the suspected defect region of the target layer and the corresponding defect-free region in the reference image. It should be noted that the reference image is an ideal defect-free image of the same batch of dies (usually the adjacent die), and the electron beam detection image is a grayscale image, each pixel contains a gray scale value, the value range is 0-255, where 0 represents pure black and 255 represents pure white, the smaller the gray scale value, the darker the point in the image, corresponding to the weaker area of electron beam signal penetration. If the image is stored in RGB format, the three channel values R, G, B of each pixel are equal, which are the gray scale value of the pixel, for example, the pixel with a gray scale value of 128 is represented as (128, 128, 128) in RGB. Therefore, the brightness distribution of the suspected defect region of the target layer and the corresponding defect-free region in the reference image can be assigned to each pixel.

[0067] Then S3 is performed, the first absolute value of the gray scale difference between each pixel of the suspected defect region and the corresponding pixel in the defect-free region is calculated, and the first average value is calculated based on the first absolute value of all pixels in the suspected defect region.

[0068] It should be noted that S4 is not performed after S3, but needs to be performed before S1 to obtain the discrimination threshold between different defect categories in advance.

[0069] In S4, determining the discrimination threshold between different defect categories based on the detection samples comprises:

[0070] For each labeled defect region in each detection sample, the second absolute value of the gray level difference between each pixel in the defect region and the corresponding pixel in the reference image is calculated, and the second average value and the standard deviation are calculated based on the second absolute values of all pixels in the defect region;

[0071] Statistical distribution analysis is performed based on the second average value and the standard deviation of the detection samples of different defect categories to obtain the probability density curve of the gray level difference distribution of each category;

[0072] The intersection point of the probability density curves of the gray level difference distributions of different categories is taken as the discrimination threshold.

[0073] By statistically analyzing a large number of artificially labeled electron beam detection samples of different products and under different detection conditions, the average value and the standard deviation of the region gray level difference of different defect categories are calculated, and then a reasonable discrimination threshold is determined.

[0074] For different defect categories, a large number of artificially labeled electron beam detection samples are first collected, and the defects include three categories of nuisance, weak voltage contrast (WVC) and dark voltage contrast (DVC). Among them:

[0075] Dark voltage contrast defect: refers to a defect in which the local gray value in the electron beam detection image is significantly lower than the surrounding normal region, usually showing a clear dark spot. This type of defect is usually caused by material structure abnormalities, incomplete CT hole or thickness abnormalities, etc., resulting in blocked electron beam penetration and obvious signal attenuation, which is a real defect that has a greater impact on product performance.

[0076] Weak voltage contrast defect: refers to a defect in which the gray level difference between the defect region and the background is small, the voltage contrast signal is weak, and the defect outline is not obvious, making detection more difficult. This type of defect usually represents a slight material abnormality or process deviation, and is an early or less severe real defect that needs to be accurately identified with the help of the method provided by the present application.

[0077] Nuisance: refers to a signal that appears in the electron beam detection image with certain gray level abnormalities, but is verified to have no substantial impact on product performance, which may be caused by noise, material surface texture or equipment fluctuations, etc. Non-defect factors. Correctly identifying nuisance defects is crucial for reducing false positives and improving detection accuracy.

[0078] For each defect area marked in the inspection sample, the statistical parameters of its grayscale difference—the second mean and standard deviation—are calculated. By performing a statistical distribution analysis on the second mean and standard deviation data for samples of different categories, using histograms and kernel density estimation methods, probability density curves for the grayscale difference distribution of each category are obtained. Subsequently, by observing the intersection of these probability density curves, the demarcation point (i.e., the discrimination threshold) between different defect categories can be determined.

[0079] For example, the intersection point of the probability density curve of the weak voltage contrast defect and the artifact is selected as the first threshold, and the intersection point of the probability density curve of the dark voltage contrast defect and the weak voltage contrast defect is selected as the second threshold.

[0080] When the first absolute value is less than a first threshold, it is determined to be an artifact;

[0081] When the first absolute value is greater than or equal to a first threshold and less than a second threshold, it is determined to be a weak voltage contrast defect;

[0082] When the first absolute value is greater than or equal to the second threshold, it is determined to be a dark voltage contrast defect.

[0083] The first and second thresholds are used to divide defect categories into three categories: dark voltage contrast defects, weak voltage contrast defects, and artifacts, which significantly improves the accuracy and robustness of defect identification. The thresholds can be dynamically adjusted in combination with subsequent yield feedback to achieve adaptive optimization.

[0084] In this embodiment, the first absolute value and the second absolute value ΔG(x, y) can both be calculated using the following formula:

[0085] ΔG(x,y)=|I EBI (x,y)-G compare (x,y)|

[0086] Among them, I EBI (x, y) is the grayscale value of the suspected defect area of ​​the target layer or the detection sample at the coordinate (x, y), G compare (x, y) is the grayscale value at the corresponding coordinate in the corresponding reference image, and |·| represents the absolute value, which is used to measure the brightness / darkness deviation.

[0087] The first average value and the second average value ΔG mean The following formula can be used for calculation:

[0088]

[0089] Where N is the total number of pixels in the defect area,

[0090] The standard deviation σ is calculated as follows:

[0091]

[0092] Then, by performing statistical distribution analysis on the second average value and standard deviation data of different categories of samples, a discrimination threshold between different defect categories can be determined.

[0093] In this embodiment, according to the first average value ΔG mean1 The size relationship between the first threshold T1 and the second threshold T2 is used for defect recognition:

[0094] If ΔG mean1 < T1, it is determined to be an artifact;

[0095] If T1≤ ΔG mean1 < T2, it is determined to be a weak voltage contrast defect;

[0096] If ΔG mean1 ≥ T2, it is determined to be a dark voltage contrast defect.

[0097] In this embodiment, the detection sample is a target layer of a defect region of different product types and the same product type under different detection conditions, so as to obtain a defect discrimination threshold suitable for different product types / detection conditions, which has strong universality and higher discrimination accuracy.

[0098] Preferably, with the continuous accumulation of production data and the dynamic change of process conditions, the discrimination threshold parameters T1 and T2 used in the present application can be continuously optimized and adaptively updated. For example, dynamic correction can be realized in the following ways:

[0099] 1) Sample-driven gray scale correction: a large number of real image samples under different products and different detection conditions are collected, and their layer structures and electrical verification results are compared to make them more consistent with actual image performance and improve the accuracy of defect determination;

[0100] 2) Defect discrimination threshold retraining: by periodic labeling and manual review, the second average value and standard deviation distribution of dark voltage contrast defects, weak voltage contrast defects and artifacts are re-counted, and the first threshold T1 and the second threshold T2 are retrained and dynamically corrected based on probability density estimation, ROC analysis and other methods, so that the defect discrimination threshold is more consistent with the real electrical performance;

[0101] 3) Adaptive adjustment mechanism: in the deployment environment, a "feedback closed loop mechanism" can be introduced, and by collecting the electrical measurement pass rate and yield fluctuation after defect determination, the matching degree of the current reference image and the threshold is verified in reverse combined with the gray scale feature, a dynamic parameter adjustment system based on multi-parameter evaluation is constructed, so as to realize the "self-learning" optimization of the gray scale discrimination strategy.

[0102] Based on the same inventive concept, the embodiments of the present application also provide a system for enhancing the detection and recognition of real defects by an electron beam, comprising:

[0103] an extraction module configured to pre-process the input electron beam inspection image to extract a target layer containing a suspected defect region;

[0104] an assignment module configured to provide a reference image matching the target layer, and assign a gray scale value to each pixel based on the suspected defect region of the target layer and the corresponding defect-free region of the reference image;

[0105] a calculation module configured to calculate a first absolute value of the gray scale difference between each pixel of the suspected defect region and the corresponding pixel of the defect-free region, and calculate a first average value based on the first absolute values of all pixels of the suspected defect region;

[0106] a threshold acquisition module configured to collect a plurality of artificially labeled inspection samples for different defect categories, and determine a discrimination threshold between different defect categories based on the inspection samples;

[0107] a discrimination module configured to output a defect discrimination result of the suspected defect region based on the size relationship between the first average value and the discrimination threshold.

[0108] Preferably, the pre-processing of the input electron beam inspection image by the extraction module comprises:

[0109] peeling off the background information of the electron beam inspection image to retain a circuit pattern region having a circuit structure;

[0110] performing layer separation on the circuit pattern region to retain the target layer containing the suspected defect region.

[0111] Preferably, the determination of the discrimination threshold between different defect categories by the threshold acquisition module based on the inspection samples comprises:

[0112] calculating a second absolute value of the gray scale difference between each pixel of the labeled defect region in each inspection sample and the corresponding pixel of the reference image, and calculating a second average value and a standard deviation based on the second absolute values of all pixels in the defect region;

[0113] statistically analyzing the second average values and the standard deviations of the inspection samples of different defect categories to obtain a probability density curve of the gray scale difference distribution of each category;

[0114] taking the intersection point of the probability density curves of the gray scale difference distributions of different categories as the discrimination threshold.

[0115] Preferably, the defect categories include dark voltage contrast defects, weak voltage contrast defects and artifacts, and the discrimination threshold includes a first threshold and a second threshold.

[0116] The output of the defect discrimination result of the target layer by the discrimination module based on the size relationship between the first absolute value and the discrimination threshold comprises:

[0117] when the first absolute value is less than the first threshold value, the artifact is determined;

[0118] when the first absolute value is greater than or equal to the first threshold value and less than the second threshold value, the weak voltage contrast defect is determined;

[0119] when the first absolute value is greater than or equal to the second threshold value, the dark voltage contrast defect is determined.

[0120] Since the system for detecting and identifying real defects by enhanced electron beams provided by the embodiments of the present application and the method for detecting and identifying real defects by enhanced electron beams provided above belong to the same inventive concept, the system for detecting and identifying real defects by enhanced electron beams provided by the present application has all the advantages of the method for detecting and identifying real defects by enhanced electron beams provided above, and thus the beneficial effects of the system for detecting and identifying real defects by enhanced electron beams provided by the present application will not be repeated here.

[0121] Based on the same inventive concept, the embodiments of the present application further provide a readable storage medium having a computer program stored thereon, and the computer program can implement the method for detecting and identifying real defects by enhanced electron beams as above when executed.

[0122] Readable storage medium can be the tangible device that can keep and store the instruction used by instruction execution device, for example, can be but not limited to electric storage device, magnetic storage device, optical storage device, electromagnetic storage device, semiconductor storage device or above-mentioned any suitable combination.The more specific example (non-exhaustive list) of readable storage medium comprises: portable computer disk, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital versatile disk (DVD), memory stick, floppy disk, mechanical coding device, for example, punch card or the convex structure in the groove that stores instruction thereon and above-mentioned any suitable combination.Computer program described herein can be downloaded to various computing / processing equipment from readable storage medium, or is downloaded to external computer or external storage device by network, for example Internet, local area network, wide area network and / or wireless network.Network can comprise copper transmission cable, optical fiber transmission, wireless transmission, router, firewall, switch, gateway computer and / or edge server. The network adapter card or network interface in each computing / processing device receives the computer program from the network and forwards the computer program for storage in a readable storage medium in the various computing / processing devices. The computer program for performing the operations of the present invention can be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The computer program can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, various aspects of the present invention are implemented by utilizing state information of a computer program to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), which can execute computer-readable program instructions.

[0123] The computer program can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions which execute via the processor of the computer or other programmable data processing apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks. The computer program can also be stored in a storage unit that can include a ROM, RAM, semiconductor chip, or other memory storage medium, etc. which can be coupled to the processor so that the computer program instructions which are read from the storage unit can implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0124] The computer program can also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus or other devices to produce a computer implemented process such that the computer program which executes on the computer, other programmable data processing apparatus, or other devices implements the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0125] Since the readable storage medium provided by the embodiment of the present application belongs to the same inventive concept as the method for enhancing electron beam detection to identify real defects, the readable storage medium provided by the present application has all the advantages of the method for enhancing electron beam detection to identify real defects, and thus the beneficial effects of the readable storage medium provided by the present application will not be repeated here.

[0126] In summary, the embodiment of the present application provides a method and system for enhancing electron beam detection to identify real defects. The target layer containing the suspected defect area is peeled off in advance, then the gray scale of the suspected defect area and the corresponding defect-free area in the reference image is assigned according to the brightness distribution, and the average value of the gray scale difference of all pixels in the suspected defect area is calculated. Finally, the average value is discriminated by using the predetermined discrimination threshold between different defect categories. The discrimination accuracy of real defects in the electron beam detection image is effectively improved, and the method is suitable for defect detection under different product models / different detection conditions, improves the detection efficiency, and further improves the yield of products.

[0127] Furthermore, it should be appreciated that the above-described embodiments are merely exemplary as to the present application and need not be practiced as described. It should also be understood that where the above-described embodiments are, for clarity, described in terms of steps or means, these steps or means need not be performed in the order given or even one at a time. Rather, certain steps or means can be performed in an order different than other steps or means or one following another, or performed at least at partially in parallel, to express the sequential order of individual steps or features or functions. Also, description of the above-described embodiments is meant to be illustrative only and not limiting as to the scope of the present application, which is set forth in the following claims.

Claims

1. A method for enhancing electron beam inspection to identify real defects, characterized in that: include: Preprocess the input electron beam inspection image to extract the target layer containing the suspected defect area; Providing a reference image that matches the target layer, and assigning a grayscale value to each pixel based on the brightness distribution of the suspected defective area of ​​the target layer and the corresponding non-defective area in the reference image; Calculating a first absolute value of a grayscale difference between each pixel in the suspected defect area and a corresponding pixel in the non-defective area, and calculating a first average value based on the first absolute values ​​of all pixels in the suspected defect area; For different defect categories, a number of manually labeled test samples are collected, and the discrimination thresholds between different defect categories are determined based on the test samples; A defect discrimination result of the suspected defect area is output based on a magnitude relationship between the first average value and the discrimination threshold.

2. The method for identifying real defects by enhanced electron beam detection according to claim 1, characterized in that: The pretreatment includes: stripping off background information of the electron beam detection image and retaining a circuit pattern area having a circuit structure; The circuit pattern area is layer-separated, and the target layer containing the suspected defect area is retained.

3. The method for identifying real defects by enhanced electron beam inspection according to claim 1, characterized in that: Determining the discrimination thresholds between different defect categories based on the detection samples includes: For each defect area marked in the inspection sample, calculate the second absolute value of the grayscale difference between each pixel contained therein and the corresponding pixel in the corresponding reference image, and calculate the second mean and standard deviation based on the second absolute values ​​of all pixels in the defect area; Perform statistical distribution analysis based on the second mean value and standard deviation of the inspection samples of different defect categories to obtain the probability density curve of the grayscale difference distribution of each category; The intersection of the probability density curves of the grayscale difference distributions of different categories is used as the discrimination threshold.

4. The method for identifying real defects by enhanced electron beam detection according to claim 1 or 3, characterized in that: The defect categories include dark voltage contrast defects, weak voltage contrast defects and artifacts, and the discrimination thresholds include a first threshold and a second threshold; Outputting the defect discrimination result of the target layer based on the magnitude relationship between the first absolute value and the discrimination threshold includes: When the first absolute value is less than the first threshold, it is determined to be the artifact; When the first absolute value is greater than or equal to the first threshold and less than the second threshold, it is determined to be the weak voltage contrast defect; When the first absolute value is greater than or equal to the second threshold, it is determined to be the dark voltage contrast defect.

5. The method for identifying real defects by enhanced electron beam inspection according to claim 1, characterized in that: The test samples are target layers of different product types and the same product type under different test conditions.

6. A system for enhancing electron beam inspection to identify real defects, characterized in that: include: An extraction module is used to pre-process the input electron beam inspection image and extract the target layer containing the suspected defect area; an assignment module, configured to provide a reference image matching the target layer and assign a grayscale value to each pixel based on the brightness distribution of the suspected defective area of ​​the target layer and the corresponding non-defective area in the reference image; a calculation module, configured to calculate a first absolute value of a grayscale difference between each pixel in the suspected defect area and a corresponding pixel in the non-defective area, and calculate a first average value based on the first absolute values ​​of all pixels in the suspected defect area; A threshold acquisition module is used to collect a number of manually labeled test samples for different defect categories and determine the discrimination thresholds between different defect categories based on the test samples; A discrimination module is configured to output a defect discrimination result of the suspected defect area based on a magnitude relationship between the first average value and the discrimination threshold.

7. The system for identifying real defects by enhanced electron beam inspection according to claim 6, characterized in that: The extraction module preprocesses the input electron beam detection image, including: stripping off background information of the electron beam detection image and retaining a circuit pattern area having a circuit structure; The circuit pattern area is layer-separated, and the target layer containing the suspected defect area is retained.

8. The system for enhancing electron beam detection and identifying real defects according to claim 6, characterized in that: The threshold acquisition module determines the discrimination thresholds between different defect categories based on the detection samples, including: For each defect area marked in the inspection sample, calculate the second absolute value of the grayscale difference between each pixel contained therein and the corresponding pixel in the corresponding reference image, and calculate the second mean and standard deviation based on the second absolute values ​​of all pixels in the defect area; Perform statistical distribution analysis based on the second mean value and standard deviation of the inspection samples of different defect categories to obtain the probability density curve of the grayscale difference distribution of each category; The intersection of the probability density curves of the grayscale difference distributions of different categories is used as the discrimination threshold.

9. The system for enhancing electron beam detection and identifying real defects according to claim 6 or 8, characterized in that: The defect categories include dark voltage contrast defects, weak voltage contrast defects and artifacts, and the discrimination thresholds include a first threshold and a second threshold; The determination module outputs a defect determination result of the target layer based on the magnitude relationship between the first absolute value and the determination threshold, including: When the first absolute value is less than the first threshold, it is determined to be the artifact; When the first absolute value is greater than or equal to the first threshold and less than the second threshold, it is determined to be the weak voltage contrast defect; When the first absolute value is greater than or equal to the second threshold, it is determined to be the dark voltage contrast defect.

10. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the method for identifying real defects by enhanced electron beam detection according to any one of claims 1 to 5 can be implemented.

Images