Wafer defect detection method, computer program product and computer equipment

Abstract

The invention provides a wafer defect detection method, a computer program product and computer equipment. The method comprises the following steps: acquiring a defect hot spot area of a detected wafer; performing electron beam scanning simulation on the defect hot spot area to obtain an electron distribution reference image of the defect hot spot area; acquiring an electron distribution detection image obtained by detecting the defect hot spot area by the electron beam detection equipment according to the simulated scanning parameters; and comparing the electron distribution reference image with the electron distribution detection image, and marking a difference position as a defect point. By means of the scheme, the number of images needing to be collected by the electron beam detection equipment is greatly reduced, the detection speed is greatly increased, and therefore the efficiency of the electron beam detection equipment is improved, and the requirement for batch manufacturing of wafers is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image data processing for semiconductor defect detection, and in particular, to a wafer defect detection method, a computer program product and a computer device. BACKGROUND

[0002] The core technical bottleneck in the field of semiconductor wafer defect detection lies in the fundamental contradiction between the atomic-level defect scale of advanced processes and the efficiency demand of mass production.

[0003] The specific performance of the above technical bottleneck is that: although optical detection technology has the advantage of high speed, it cannot analyze the nanoscale defect morphology due to the optical diffraction limit, and can only provide defect coordinates, and the false positive rate for multi-layer stacked structure is more than 40%; although electron beam detection technology can realize atomic-level resolution and three-dimensional defect reconstruction, it is limited by the single-point scanning mechanism and beam stability requirements, and the detection speed is extremely low, which leads to the production bottleneck in the hybrid detection mode. The combination scheme of optical detection for preliminary screening and electron beam detection for point-by-point review has appeared in the prior art. The optical detection device performs defect preliminary screening on the wafer under test, records the defect information such as position and shape, the electron beam detection device performs coordinate calibration according to the defect information, and performs point-by-point image capture according to the defect information to perform defect re-inspection and obtain the defect re-inspection result. Due to the time-consuming of repeated positioning and the interference of optical artifacts, the scheme cannot meet the economic requirements of advanced process defect coverage and online full inspection, which seriously restricts the yield climbing.

[0004] The above technical problems reflect the irreconcilable conflict between detection accuracy and detection speed: optical detection is limited by physical laws to approach the technical limit, while the efficiency bottleneck of electron beam detection makes it long-term limited to research and development and sampling inspection scenes. Therefore, the field urgently needs breakthrough innovation to realize the unity of atomic-level precision and mass production efficiency of semiconductor processes. SUMMARY

[0005] In view of the above problems, the present application provides a wafer defect detection method, a computer program product and a computer device which overcome the above problems or at least partially solve the above problems.

[0006] An object of the present application is to reduce the number of image capture of the electron beam detection device and improve the image capture speed.

[0007] In particular, the present application provides a wafer defect detection method, comprising: obtaining a defect hotspot area of a wafer under test; performing electron beam scanning simulation on the defect hotspot area to obtain an electron distribution reference image of the defect hotspot area; obtaining an electron distribution detection image obtained by the electron beam detection device detecting the defect hotspot area according to the simulated scanning parameters; The electron distribution reference image and the electron distribution detection image are compared, and the difference position is marked as a defect point.

[0008] Optionally, the step of comparing the electron distribution reference image and the electron distribution detection image comprises: An image processing model is obtained in advance; The electron distribution reference image and the electron distribution detection image are processed into quality-enhanced images under the same parameter reference by using the image processing model, forming a pair of comparison images; The pair of comparison images are used for difference comparison.

[0009] Optionally, the step of constructing the image processing model comprises: A source model for image quality enhancement processing is obtained; The source model is subjected to transfer learning using a pre-acquired electron beam scanning image dataset, and the image processing model is constructed, wherein the images in the electron beam scanning image dataset are pre-labeled according to whether they meet the required conditions for comparison.

[0010] Optionally, the parameter reference comprises any one or more of the following: image contrast reference, image density reference, image sharpness reference.

[0011] Optionally, the step of obtaining the defect hotspot area of the wafer under test comprises: An optical proximity correction process of the wafer under test is performed, and a hotspot position found by the optical proximity correction process is recorded; The hotspot position is converted into a defect hotspot area by using an offline recipe editing function in the optical proximity correction process, and the defect hotspot area is recorded in a process parameter combination file generated by the optical proximity correction process.

[0012] Optionally, after the step of converting the hotspot position into the defect hotspot area, the method further comprises: configuring a parameter set of the probe die according to the defect hotspot area, and keeping the parameters of the reference die blank and unconfigured; the parameter set of the probe die comprises: index, coordinates, and size of the die; and After obtaining the electron distribution reference image of the defect hotspot area, the method further comprises: storing the parameter set of the probe die and the electron distribution reference image in the process parameter combination file, respectively.

[0013] Optionally, the step of obtaining the electron distribution detection image obtained by the electron beam detection device in the defect hotspot area according to the simulated scanning parameters comprises: The process parameter combination file is read, and the defect hotspot area is converted into detection parameters of the electron beam detection device; The defect hotspot area is detected by the electron beam detection device according to the detection parameters and the scanning parameters; An image detected by the electron beam detection device is taken as an electron distribution detection image and stored.

[0014] Optionally, the scanning parameters include a scanning point average parameter, a scanning line average parameter, and a scanning frame average parameter.

[0015] According to still another aspect of the present application, there is also provided a computer device comprising a memory, a processor, and a machine executable program stored in the memory and running on the processor, and the processor implements the steps of any of the wafer defect detection methods when executing the machine executable program.

[0016] The wafer defect detection method of the present application acquires a defect hotspot area of a wafer to be detected, obtains an electron distribution reference image by using electron beam scanning simulation, detects an electron distribution detection image by an electron beam detection device in the defect hotspot area according to the simulated scanning parameters, processes the electron distribution reference image and the electron distribution detection image into a contrast image pair satisfying a difference contrast condition, and can accurately mark the difference position as a defect point, thereby improving the pertinence and accuracy of wafer defect detection, reducing unnecessary detection procedures compared with comprehensive detection, and improving detection efficiency.

[0017] Further, the wafer defect detection method of the present application is based on an AI image enhancement algorithm, constructs a small-scale Scanning Electron Microscope (SEM) image dataset, fine-tunes a model using data migration technology, constructs an image processing model, and is used to realize Image Quality Enhancement (IQE), improve the signal-to-noise ratio and image sharpness, adjust the image contrast and image density parameters of the image pair to a unified benchmark, and improve the accuracy of image difference contrast.

[0018] Still further, the wafer defect detection method of the present application utilizes the hotspot query function of the Optical Proximity Correction (OPC) process to obtain a defect hotspot area, converts the defect hotspot area into detection parameters of the electron beam detection device when detecting by the electron beam detection device, detects in the defect hotspot area by the electron beam detection device according to the detection parameters and scanning parameters, avoids using a Graphic Data System (GDS) to design the defect hotspot area, avoids the difficulty of hotspot positioning by the electron beam detection device, and reduces invalid detection.

[0019] The above and other objects, advantages and features of the present application will become more apparent after a reading of the following detailed description of the embodiments thereof, taken in conjunction with the annexed drawings. BRIEF DESCRIPTION OF DRAWINGS

[0020] Some specific embodiments of the present application will be described in detail in the following with reference to the attached drawings. The same or similar components or parts are designated by the same reference numerals, and a repeated description thereof will be omitted. It should be understood by those skilled in the art that the drawings are not necessarily drawn to scale. In the drawings: Figure 1 is a schematic diagram of a wafer defect detection method according to an embodiment of the present application; Figure 2 is a schematic diagram of an image comparison step in a wafer defect detection method according to an embodiment of the present application; Figure 3 is a schematic diagram of a construction step of an image processing model in a wafer defect detection method according to an embodiment of the present application; Figure 4 is a schematic diagram of an OPC flow in a wafer defect detection method according to an embodiment of the present application; Figure 5 is a schematic diagram of a detection flow of an electron beam detection device in a wafer defect detection method according to an embodiment of the present application; Figure 6 is a schematic diagram of a computer program product according to an embodiment of the present application; Figure 7 is a schematic diagram of a computer readable storage medium according to an embodiment of the present application; Figure 8 is a schematic diagram of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] It should be understood by those skilled in the art that the embodiments described below are only a part of the embodiments of the present application, rather than all the embodiments of the present application, and are intended to explain the technical principles of the present application, rather than to limit the protection scope of the present application. Based on the embodiments provided by the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor should fall within the protection scope of the present application.

[0022] It is to be appreciated that the logic and / or steps represented in the flow diagrams and / or described herein, for example, can be embodied in non-transitory computer- readable medium, which can be executed by an instruction execution system, apparatus, or device; such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions; or combination thereof.

[0023] In the semiconductor production process, the resolution limit of optical imaging itself and the design problem of mask pattern can cause potential defects on the wafer, and the positions of the potential defects can be referred to as hot spots, and the areas where the hot spots are located are referred to as defect hot spot areas.

[0024] Optical proximity correction (OPC) is a technical means for compensating for optical proximity effect (OPE) in the photolithography process. OPC adds some auxiliary patterns or changes the shape, size, position, etc. of the original pattern on the mask to offset the influence of the optical proximity effect, so that the pattern formed on the wafer is as close as possible to the ideal pattern designed, and the precision and yield of chip manufacturing are improved. In the OPC process, a hot spot finding function (HotSpot Finder) can be added, which determines the position of the Hot Spot on the wafer in the relative position by combining the material and geometric information of the previous layer and the current layer of the chip.

[0025] In some related technologies of the present application, a script can be developed according to GDS and lithography exposure effect design rules, hot spots are generated on the GDS using the script, and then a clustering technique is used to convert the hot spots into defect detection hot spot areas (or care areas). The hot spot area is further converted into an electron beam detection device detection parameter (Recipe). The electron beam detection device performs online detection according to the detection parameter. The detection parameter can include detection position, scanning mode, etc.

[0026] However, generating hot spots on the GDS using a script requires a full understanding of the GDS design and the lithography exposure performance, which makes it difficult to accurately locate the hot spots. On the other hand, the electron beam detection device detects according to the detection parameter corresponding to the generated hot spot, and the throughput still cannot meet the needs of the wafer factory.

[0027] The wafer defect detection method of the embodiment can accurately predict a defect hotspot area, and reduce the number of SEM images required for detection by reducing the detection conditions of the electron beam detection device, thereby improving detection efficiency and reducing detection time.

[0028] Figure 1 FIG. 1 is a schematic diagram of a wafer defect detection method according to an embodiment of the present application, which can generally include: In step S101, a defect hotspot area of a wafer under test is obtained.

[0029] The defect hotspot area can be determined during the execution of an optical proximity correction process. For example, the step of obtaining the defect hotspot area of the wafer under test can include: executing an optical proximity correction process of the wafer under test, and recording hotspot positions found by the optical proximity correction process; converting the hotspot positions to a defect hotspot area using an offline recipe editing function in the optical proximity correction process, and recording the defect hotspot area in a process parameter combination file generated by the optical proximity correction process. Thus, this step can accurately obtain the defect hotspot area based on the optical proximity correction process, provide an accurate detection range for subsequent targeted detection, and improve the accuracy and targeting of the detection. The process parameter combination file Recipe can be used to record a parameter set of a detection die and an electron distribution reference image obtained by simulation in the next step.

[0030] In step S102, electron beam scanning simulation is performed on the defect hotspot area to obtain an electron distribution reference image of the defect hotspot area. The electron distribution reference image obtained by electron beam scanning simulation serves as a reference image (golden image) for subsequent comparison, and the electron distribution reference image represents an ideal state of the semiconductor.

[0031] In step S103, an electron distribution detection image obtained by detection of the defect hotspot area by the electron beam detection device according to the simulated scanning parameters is obtained.

[0032] The electron beam detection device can obtain relevant information of the defect hotspot area according to the process parameter combination file (Recipe) generated by the optical proximity correction process in advance. In some embodiments, the optical proximity correction process can further configure a parameter set of a detection die according to the defect hotspot area after the step of converting the hotspot positions to the defect hotspot area, and keep the parameters of a reference die blank and not configured. The parameters of the reference die are replaced by the electron distribution reference image.

[0033] The parameter set of the detection die and the electron distribution reference image are respectively stored in the process parameter combination file Recipe. The process parameter combination file Recipe is exported and used for detection by the electron beam detection device and image difference comparison.

[0034] The parameter set of the detected die can include index, coordinates, and size of the die, which define the position of the defect hotspot region, thereby greatly reducing the scanning range of the electron beam detection device and improving the efficiency.

[0035] The detecting step performed by the electron beam detection device can include: reading the process parameter combination file and converting the defect hotspot region into detection parameters of the electron beam detection device; detecting, by the electron beam detection device, the defect hotspot region according to the detection parameters and scanning parameters; and storing the image detected by the electron beam detection device as an electron distribution detection image. The detection process of the electron beam detection device is more targeted, thereby shortening the scanning time of the wafer.

[0036] The scanning parameters include dot average parameters, line average parameters, and frame average parameters. In the embodiment, each defect hotspot region is simulated to obtain the same scanning physical effects of the electron gun in the defect hotspot region in terms of dot average, line average, and frame average. The electron beam detection device also performs the same electron beam scanning in the defect hotspot region in terms of dot average, line average, and frame average. The dot average, line average, and frame average can be set to 1, and the advantages of the simulation algorithm are utilized to improve the speed of image acquisition, which can ensure stable and reliable scanning image quality and improve the speed of image acquisition.

[0037] In step S104, the electron distribution reference image and the electron distribution detection image are compared, and the difference position is marked as a defect point.

[0038] Figure 2 FIG. 1 is a schematic diagram of the image comparison step in the wafer defect detection method according to an embodiment of the present application. The step of comparing the electron distribution reference image and the electron distribution detection image can include: In step S201, an image processing model is obtained. In step S202, the electron distribution reference image and the electron distribution detection image are processed into quality-enhanced images under the same parameter reference by using the image processing model, to form a pair of comparison images. In step S203, the pair of comparison images are compared for differences.

[0039] The electron distribution detection image is obtained by scanning a real wafer by an electron beam detection device, which may be affected by scanning conditions, wafer material characteristics, device noise, etc., thereby having differences in image quality from the electron distribution reference image, which may cause misjudgment in comparison. Therefore, the method of the embodiment adjusts the image quality to a uniform level through image quality enhancement to perform difference comparison on the comparison image pair.

[0040] The processing process of the comparison image pair is realized by artificial intelligence (AI) image processing. After processing, the image quality of the comparison image pair is greatly improved, comparison can be performed under the same standard, comparison errors caused by differences in image quality parameters are eliminated, and the accuracy and reliability of the defect detection result are further improved.

[0041] Figure 3 FIG. 1 is a schematic diagram of a construction step of an image processing model in a wafer defect detection method according to an embodiment of the present application. The construction step of the image processing model can include: In step S301, a source model for image quality enhancement processing is obtained. In step S302, the source model is subjected to transfer learning using a pre-acquired electron beam scanning image dataset to construct an image processing model. The images in the electron beam scanning image dataset are pre-labeled according to whether they meet the required conditions for comparison. Using the labeled electron beam scanning image dataset to perform transfer learning to construct the image processing model can make the model more suitable for the characteristics of the electron beam scanning image and reduce the amount of data required for model training. Using a general image processing model, the electron beam scanning image dataset is fine-tuned to better adapt to the application scenario of defect recognition in the embodiment.

[0042] The parameter reference includes any one or more of the following: image contrast reference, image density reference, and image sharpness reference. Image contrast reflects the light and dark contrast between different regions in the image (such as defects and normal regions, material boundaries, etc.). In a high-contrast image, the target region and the background differ significantly, and the defect outline is easier to identify; in a low-contrast image, the details may be lost due to blurred light and dark differences. Image density reflects the signal intensity distribution density of pixels in the image, and too dark or signal fluctuation affects the accuracy of comparison. Image sharpness reflects the clarity of edges and richness of details in the image. In a high-sharpness image, the edges are clear and the details are rich. The uniformity of the parameter reference provides more detailed and accurate parameter specifications for finding defect points in accurate comparison images, improving the accuracy of comparison. The image contrast reference, image density reference, and image sharpness reference are respectively configured according to the characteristics of the wafer image and the comparison needs.

[0043] The wafer defect detection method of the embodiment acquires a defect hotspot area of a wafer to be detected, obtains an electron distribution reference image by using electron beam scanning simulation, detects an electron distribution detection image in the defect hotspot area according to the simulated scanning parameters by an electron beam detection device, and processes the electron distribution reference image and the electron distribution detection image into a contrast image pair meeting a difference contrast condition, so that a difference position can be accurately marked as a defect point, the wafer defect detection pertinence and accuracy are improved, unnecessary detection processes are reduced compared with comprehensive detection, and the detection efficiency is improved. The number of images that need to be collected by the electron beam detection device is greatly reduced, the detection speed is greatly improved, the efficiency of the electron beam detection device is improved, and the wafer batch manufacturing demand is met.

[0044] The method of the embodiment mainly includes two parts, the first part is improvement of an OPC process to determine a defect hotspot area, and the second part is modification of an inspection process of an electron beam detection device.

[0045] Figure 4 FIG. 1 is a schematic diagram of an OPC process in a wafer defect detection method according to an embodiment of the present application; the OPC process can include: Step S401, performing an optical proximity correction OPC process; Step S402, running a hot spot finder function, combining material and geometric information of a previous layer and a current layer of a chip and a relative position on a wafer to determine a hot spot position; Step S403, using an offline recipe editor function to convert the hot spot position into a suitable defect hotspot area care area.

[0046] Step S404, setting reference die information, setting only a detection die, and keeping parameters of the reference die blank and not configured; Step S405, setting other parameters of a process parameter combination file recipe, archiving and exporting.

[0047] Figure 5 FIG. 2 is a schematic diagram of an inspection process of an electron beam detection device in a wafer defect detection method according to an embodiment of the present application; the inspection process of the electron beam detection device can include: Step S501, for each defect hot spot area, perform electron beam scanning simulation, simulate the scanning physical effect of the electron gun in the Hot Spot area, and perform dot average, line average, and frame average all equal to 1.

[0048] Step S502, collect the electron distribution image (Electron distribution image, EDI) formed by the secondary electrons in the defect hot spot area Hot Spot area as the electron distribution reference image golden image, and store it. The defect detection algorithm is set to be based on the golden image inspection process, which can reduce the number of image captures.

[0049] Step S503, store the electron distribution reference image golden image into the process parameter combination file Recipe reference image.

[0050] Step S504, the electron beam detection device reads the process parameter combination file Recipe, and detects the defect hot spot area with scanning parameters of dot average, line average, and frame average all equal to 1, and captures the electron distribution detection image.

[0051] Step S505, the electron distribution detection image and the corresponding electron distribution reference image golden image form a comparison image pair, and perform defect detection calculation.

[0052] Step S506, process the electron distribution detection image and the corresponding electron distribution reference image through the image quality enhancement algorithm IQE, improve the signal-to-noise ratio and image sharpness, and adjust the image contrast and image density of the image pair to a unified level. The image quality enhancement algorithm IQE can be an AI-based image enhancement algorithm. By constructing a small-scale SEM image dataset and using data migration technology to fine-tune the model, the adaptability of the model is improved.

[0053] Step S507, the IQE-processed image pair is used for defect detection, the electron distribution detection image is compared with the electron distribution reference image, and the places where the images are different are marked to identify the defects and their positions.

[0054] The method breaks through the technical barrier through the OPC technology, identifies the hot spot, greatly improves the detection efficiency of the electron beam detection equipment by reducing the unnecessary defect hot spot area, and through the image simulation and image enhancement technology based on AI, the golden image-based inspection process can be carried out under the lowest sampling condition, so as to further improve the sampling speed and greatly improve the equipment throughput.

[0055] The embodiment also provides a computer program product 10, a computer readable storage medium 20, and a computer device 30. Figure 6 is a schematic diagram of the computer program product 10 according to an embodiment of the present application, Figure 7 is a schematic diagram of the computer readable storage medium 20 according to an embodiment of the present application, Figure 8 is a schematic diagram of the computer device 30 according to an embodiment of the present application. The computer program product 10 comprises a computer program 11, which, when executed by the processor 32, implements the steps of any of the wafer defect detection methods described above. The computer readable storage medium 20 has the computer program 11 stored thereon, and the computer program 11, when executed by the processor 32, implements the steps of any of the wafer defect detection methods described above. The computer device 30 can comprise a memory 31, a processor 32, and a computer program 11 stored on the memory 31 and running on the processor 32.

[0056] Computer program 11, which can also comprise programs 11 for performing operations of the present application, can be assembly instructions, Instruction Set Architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state setting data, configuration data for an integrated circuit, or source code or object code written in any combination of one or more programming languages and processes. Computer program 11 can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the latter scenario, 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 the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate array (FPGA), or programmable logic array (PLA) can execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present application.

[0057] For the purposes of this description of the embodiments, computer program product 10 is a product of authorship that comprises computer program 11. For the purposes of this description of the embodiments, computer readable storage medium 20 is a tangible device that can retain and store computer program 11 for purposes of having computer program 11 accessible by a processing device or machine or combination of both. More specifically, computer readable storage medium 20 can be, for example, but is not limited to, portable computer diskette, 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, mechanically encoded device such as punch-cards or raised structures in a groove having information recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium 20, as used herein, is intended to be broadly construed to encompass a tangible device that can retain and store instructions for execution by a processing device or machine or combination of both.

[0058] The computer device 30 can be, for example, a server, a desktop computer, a notebook computer, a tablet computer, or a smart phone. In some examples, the computer device 30 can be a cloud computing node. The computer device 30 can be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules can include routines, programs, objects, components, logic, data structures, and so on that perform particular tasks or implement particular abstract data types. The computer device 30 can be practiced in distributed cloud computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules can be located in both local and remote computer system storage media including memory storage devices.

[0059] The computer device 30 can include a processor 32 adapted to execute stored instructions, a memory 31 that provides temporary storage for operations of the instructions during operation. The processor 32 can be a single core processor, a multi-core processor, a computing cluster, or any number of other configurations. The memory 31 can include random access memory (RAM), read only memory, flash memory, or any other suitable storage system.

[0060] The computer device 30 can also include a network adapter / interface and an input / output (I / O) interface. The I / O interface allows for input and output of data with external devices that can be connected to the computer device. The network adapter / interface can provide for communication between the computer device and a network, typically illustrated as a communication network.

[0061] To this end, it will be appreciated that, although specific embodiments of the application have been described herein for purposes of illustration, various modifications can be made in the details of the application without departing from the spirit and scope thereof. Therefore, it is to be understood that the scope of the application is not to be limited to the specific examples disclosed and that modifications or variations are possible so long as the spirit or principles of the application remain unaltered.

Claims

1. A wafer defect detection method, characterized in that include: Obtain the defect hotspot area of ​​the inspected wafer; Performing electron beam scanning simulation on the defect hotspot area to obtain an electron distribution reference image of the defect hotspot area; Acquire an electron distribution detection image obtained by an electron beam detection device detecting the defect hotspot area according to the simulated scanning parameters; The electron distribution reference image and the electron distribution detection image are compared, and the difference positions are marked as defect points.

2. The wafer defect detection method according to claim 1, wherein: The step of comparing the electron distribution reference image and the electron distribution detection image comprises: Get pre-built image processing models; Processing the electron distribution reference image and the electron distribution detection image into quality-enhanced images under the same parameter benchmark using the image processing model to form a comparison image pair; The compared image pairs are used to perform difference comparison.

3. The wafer defect detection method according to claim 2, wherein: The steps of constructing the image processing model include: obtaining a source model for image quality enhancement processing; The image processing model is constructed by performing transfer learning on the source model using a pre-collected electron beam scanning image dataset, wherein the images in the electron beam scanning image dataset are pre-labeled based on whether they meet the conditions required for comparison.

4. The wafer defect detection method according to claim 2, wherein: The parameter benchmark includes any one or more of the following: image contrast benchmark, image density benchmark, and image sharpness benchmark.

5. The wafer defect detection method according to claim 1, wherein: The step of obtaining the defect hotspot area of ​​the inspected wafer includes: executing an optical proximity correction process for the inspected wafer and recording a hotspot position found by the optical proximity correction process; The hotspot position is converted into the defect hotspot area by using the offline recipe editing function in the optical proximity correction process, and is recorded in a process parameter combination file generated by the optical proximity correction process.

6. The wafer defect detection method according to claim 5, wherein: After the step of converting the hotspot position into the defect hotspot area, the method further includes: configuring a parameter set for detecting the grain according to the defect hotspot area, and keeping the parameters of the reference grain blank without configuration; the parameter set for detecting the grain includes: the index, coordinates, and size of the grain, and After obtaining the electron distribution reference image of the defect hotspot area, the method further includes: storing the parameter set of the detected grain and the electron distribution reference image in the process parameter combination file respectively.

7. The wafer defect detection method according to claim 5, wherein: The step of obtaining an electron distribution detection image obtained by an electron beam detection device detecting the defect hotspot area according to the simulated scanning parameters includes: Reading the process parameter combination file and converting the defect hotspot area into detection parameters of the electron beam detection equipment; The electron beam detection device detects the defect hotspot area according to the detection parameters and the scanning parameters; The image detected by the electron beam detection device is used as the electron distribution detection image and stored.

8. The wafer defect detection method according to claim 1, wherein: The scanning parameters include scanning point average parameters, scanning line average parameters, and scanning frame average parameters.

9. A computer program product, characterized in that The invention comprises a computer program, which implements the steps of the wafer defect detection method according to any one of claims 1 to 8 when the computer program is executed by a processor.

10. A computer device, characterized in that The device comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of the wafer defect detection method according to any one of claims 1 to 8 are implemented.

Images