Method for filling two-dimensional grain boundary deficiency based on three-dimensional information

Abstract

According to the two-dimensional grain boundary filling method based on the three-dimensional information, firstly, an image is collected through a high-resolution two-dimensional imaging technology, reference registration is carried out, and accurate association of spatial information is ensured; after a first-layer metallographic pattern grain boundary is completed through hand drawing, preprocessing operations such as graying, enhancing and filtering are carried out on an image. Noisy points are removed by using a specific function and a formula, binarization is carried out, and a grain boundary is refined by using professional software. And performing grain corrosion by setting a threshold value to obtain a central mark point, and calculating a proportion parameter of adjacent grains by eight-direction tracking. And detecting and scaling the crystal grain edge of the first-layer metallographic pattern, and filling the second-layer crystal grain by using a special program until a complete metallographic pattern is output. According to the method, a filling result is verified by combining a three-dimensional reconstruction technology, and an algorithm is optimized. According to the technology, the structural integrity and performance of the crystal material are effectively improved, the grain boundary reconstruction precision is improved, the material performance prediction reliability is enhanced, the experiment and calculation cost is reduced, the data processing efficiency is improved, the technology is suitable for various materials, and a new way is developed for material microcosmic research and application.

Description

Technical Field

[0001] The present invention relates to the field of material science, and in particular to a method for filling two-dimensional grain boundary defects based on three-dimensional information. Background Art

[0002] In the field of materials science, obtaining metallographic images of metal samples is a common laboratory testing method. Grain boundaries are the interface between adjacent grains and are the most common lattice defects in materials. Microscopic grain size, grain boundaries and organizational structure provide important basis for the mechanical properties of materials, such as tensile strength, toughness, and plasticity, and are therefore widely used in various design and manufacturing fields.

[0003] The morphology of grains in metallographic images can be used to predict the relevant properties of materials. The two-dimensional image of grains occupies a core position in the study of organizational structure and structural properties, and is the key to a deep understanding of the microscopic characteristics of materials. Grain boundaries are usually obscured by volume, and their extraction process is extremely challenging. In particular, the extraction of grain boundaries and the connection and reconstruction of missing grain boundaries have always been the core difficulty and focus in metallographic image processing. Accurate extraction and reconstruction of grain boundaries is one of the key technologies in the field of metallographic analysis, and the accuracy of the results directly affects the accuracy of the entire analysis process. In the process of grain boundary extraction and reconstruction, the core element is to implement an efficient, accurate and reasonable segmentation strategy for the image. In the field of metallographic analysis, professionals mainly rely on careful observation and accurate recording of key parameters such as grain size through microscopes. However, given the inherent instability of manual operation, including but not limited to time consumption, experience differences and fluctuations in technical level, these factors may significantly expand the error range of the analysis results. Because the accuracy and reliability of the analysis process have a great impact on the accuracy of grain boundary extraction and reconstruction. Therefore, based on the deep integration of computer technology and image processing technology, this paper innovatively proposes a 3D thinking algorithm to solve the problem of filling missing grain boundaries in metallographic image processing. The microstructure of metal materials has a decisive influence on their mechanical properties, and the microstructure of metallographic images is an important basis for analyzing these properties. In the preparation process of metal sample images, defects such as blurred or missing grain boundaries are inevitable, which bring significant challenges to grain boundary extraction. In view of the inefficiency of manual processing of complex metallographic images, it is very necessary to automatically fill missing grain boundaries.

[0004] In the existing technology, some progress has been made in grain boundary analysis using three-dimensional information (such as three-dimensional electron microscope images of crystals, X-ray tomography, etc.), but how to accurately fill the two-dimensional grain boundary defects based on these three-dimensional information is still a technical problem. Therefore, a method for filling the two-dimensional grain boundary defects based on three-dimensional information is proposed, in order to achieve efficient and accurate grain boundary repair. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art and in order to effectively improve the structural integrity and performance of crystal materials, the present invention provides a method for filling two-dimensional grain boundary missing parts based on three-dimensional information. The method can make full use of the concept of spatial position to fill the two-dimensional missing grain boundary map, thereby realizing an efficient, complete and accurate two-dimensional grain boundary map and facilitating subsequent three-dimensional reconstruction.

[0006] In order to solve the above technical problems, this application provides the following technical solutions: A series of metallographic images of the sample are collected using high-resolution two-dimensional imaging technology, and all metallographic images are benchmarked and registered based on specific markers using an image registration algorithm to accurately associate spatial information, wherein the image registration algorithm uses specific instructions of professional image processing software to achieve precise positioning; The grain boundaries of the first layer of metallographic images are hand-painted to make them complete, and then all metallographic images are preprocessed. The preprocessing operations include grayscale processing using floating-point calculation method, image enhancement based on equalization technology formula, and image smoothing filtering combined with Gaussian filtering and wavelet transform to ensure that the image quality meets the requirements of subsequent processing; Use specific functions to eliminate noise in metallographic images, perform binarization operations according to predetermined formulas, and then use professional software tools to refine problems such as inconsistent grain boundary width and burrs to obtain clear grain boundary images; By setting different threshold disks, multi-scale iterative corrosion operations are performed on each grain in the metallographic image while maintaining the geometric characteristics of the grain boundary. The center points of the grains are obtained at a specific iteration state before the grains completely disappear, and these center points are recorded as seed record mark points to provide a basis for subsequent analysis. For each seed record mark point, eight azimuth tracking strategies with 45° intervals are implemented, starting from the mark point and continuously tracking until the grain boundary is encountered. In this process, the length of each line segment is accurately recorded, and then the ratio parameters of adjacent grains are calculated to determine the relative relationship between grains; Perform edge detection and contour extraction on each grain in the first layer of the complete metallographic image, and perform scaling operations based on the proportional parameters calculated between two adjacent layers to provide an accurate template for subsequent filling; A specially developed program is used to fill each grain of the second layer metallographic diagram in turn. During the filling process, intelligent judgment and operation are performed based on the results of previous calculations and processing, and the cycle is repeated until all grains of the layer are filled to obtain a complete metallographic diagram of the second layer. The above filling process is executed cyclically, and the metallographic images of subsequent layers are processed in turn until the complete metallographic images of all layers are output, so as to realize the comprehensive automatic filling of grain boundaries and ensure the accuracy and consistency of the filling results.

[0007] The present invention provides a method for filling two-dimensional grain boundary defects based on three-dimensional information, comprising: Step 1: read in a series of metallographic image stacks of the sample, and perform benchmark registration on all metallographic images to associate spatial information; Step 2: Hand-draw the first layer of complete grain boundary metallographic diagram and pre-process all metallographic diagrams; Step 3: Remove noise points and perform binarization before performing refinement processing; Step 4: Iteratively etch each grain in the metallographic image to find the grain center and obtain multiple seed record mark points; Step 5: Perform eight-direction tracking on each grain to calculate the proportional parameters of its adjacent layers; Step 6: Perform edge detection and extraction on the first layer of the complete metallographic image and scale it accordingly; Step 7: Fill each grain of the second layer metallographic image in turn to obtain a complete metallographic image of the second layer; Step 8: Obtain complete metallographic images of all layers in sequence and output them.

[0008] Preferably, the S1 includes: S1.1 After heat treatment, the sample is inlaid, ground, marked with diamond points, polished, etched, photographed, etc. to obtain the first metallographic image. The polishing, etching, photographing operations are repeated to obtain the second, third, and other series of metallographic images of the same area.

[0009] S1.2 In the process of processing specific areas, image j software is used to perform rigorous image processing operations on a series of metallographic images. Specifically, the diamond-shaped markers are used as references, and plugins, registration, and SIFT commands are used to ensure that all images are accurately registered in the upper and lower positions to achieve consistency and accuracy in image analysis. A series of image stacks can be generated by accurately cropping the required area.

[0010] Preferably, S2 includes: S2.1 Manually improve the first metallographic image to ensure the integrity of its grain boundaries. The grain boundary filling work of all subsequent metallographic images can be completed automatically by computer; S2.2 grayscales all metallographic images, performs image enhancement and image smoothing filtering; S2.2.1 Grayscale processing uses floating point calculation method; S2.2.2 Image enhancement uses equalization technology formula; S2.2.3 Image smoothing filtering uses Gaussian filtering and wavelet transform.

[0011] Preferably, S3 includes: S3.1 uses the imfill function to eliminate the noise in the metallographic image; S3.2 performs binarization operation on the metallographic image using the formula; S3.3 uses MATLAB to refine the undesirable phenomena such as inconsistent grain boundary width and burrs.

[0012] Preferably, the specific steps required in step S4 are: S4.1 sets different thresholds disk, and performs multi-scale iterative corrosion processing on the given image while maintaining the geometric features of the grain boundary; in this process, each time the corrosion operation is completed, the boundary of the grain will shrink one layer inward until the grain is in an iterative state before it completely disappears, and the center point of the grain is obtained; S4.2 uses the mathematical model Gaussian curvature or mean curvature to describe the grain boundary curvature, calculates the degree of curvature through the local grain boundary geometric characteristics, and infers the shape of the filled area and the surface structure.

[0013] S4.2.1 The curvature is calculated through the local grain boundary tangent plane equation and numerical methods, and the shape of the filled area is dynamically adjusted. The crystal symmetry and space group information are combined to ensure that the lattice structure of the filled area is consistent with the surrounding area.

[0014] S4.2.2 Use clustering algorithms to cluster crystal structure regions and analyze the similarities and differences between clusters. Combine clustering results with physical properties to build a data-driven similarity assessment framework. By adjusting the similarity threshold, accurately predict or fill in the characteristics of similar grains.

[0015] S4.2.3 For crystal structures, similarity metrics such as geometric distance, phase similarity, and energy minimization can be used to assign similarity scores to each grain to influence the filling strategy. Through a dynamic adjustment mechanism, the filling threshold and similarity criteria are adjusted in real time based on the grain distribution characteristics. Incremental learning is used to optimize parameter configuration as data accumulates and the algorithm iterates.

[0016] S4.3 records the center point of each grain before it disappears, and obtains multiple seed recording mark points.

[0017] Preferably, S5 includes: S5.1 implements a tracking strategy of eight positions (45° apart) for each marker point; S5.2 Ensure that the line is traced continuously from the marked point until the grain boundary is encountered, and the length of each line segment is recorded and the proportional parameter is calculated.

[0018] Preferably, S6 includes: S6.1 performs edge detection and contour extraction on each grain in the first layer of the complete metallographic image; S6.2 is scaled according to the ratio parameters of two adjacent layers.

[0019] Preferably, the S7 includes: S7.1 Use the program to fill a die; S7.2 calculates and selects the optimal Boundary IoU metric to evaluate the matching degree between the object boundary in the segmentation result and the true boundary and measure the boundary accuracy.

[0020] S7.3 loop fills all the grains of a layer of metallographic image.

[0021] Preferably, after obtaining and outputting the complete metallographic images of all layers in sequence in step 8, a step of verifying the filling results is also included. The verification method uses three-dimensional reconstruction technology to ensure that the filled grain boundaries meet the overall structure and performance requirements of the crystal. If the verification results do not meet the requirements, the filling algorithm is dynamically optimized and adjusted based on the verification feedback.

[0022] Compared with the prior art, the method for filling two-dimensional grain boundary defects based on three-dimensional information of the present invention has at least the following beneficial effects: 1. Improved accuracy of grain boundary reconstruction: Traditional two-dimensional grain boundary analysis may have problems with missing or incomplete information, while the present invention can more accurately restore the integrity of grain boundaries by combining three-dimensional information, thus avoiding the limitations of two-dimensional data.

[0023] 2. Improved the reliability of material performance prediction: Grain boundaries play a vital role in material performance. After filling the missing grain boundaries with this method, the structural characteristics of the material can be more realistically reflected, thus providing a more accurate basis for the prediction and optimization of material performance.

[0024] 3. Reduced experimental and computational costs: Through effective three-dimensional information filling, the number of actual grain boundary experiments required is reduced, reducing experimental costs. At the same time, the efficiency of simulation and calculation is optimized due to the improved integrity of data during the calculation process.

[0025] 4. Improved data processing efficiency: This method can automatically process a large amount of crystal structure data, reduce the time for manual intervention and data processing, and improve scientific research efficiency.

[0026] 5. The present invention has wide applicability: The filling method of the present invention is not only applicable to metal materials, but can also be extended to the grain boundary reconstruction of different materials such as ceramics and semiconductors, and has strong applicability and versatility.

[0027] The following is a further description of a method for filling two-dimensional grain boundary defects based on three-dimensional information according to the present invention in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1: A flow chart of a method for filling two-dimensional grain boundary defects based on three-dimensional information according to the present invention; Figure 2 : A series of pictures of a method for filling two-dimensional grain boundary defects based on three-dimensional information according to the present invention; Figure 3 : This is a histogram equalized image of a method for filling two-dimensional grain boundary gaps based on three-dimensional information according to the present invention; Figure 4 : The present invention discloses a method for filling two-dimensional grain boundary defects based on three-dimensional information, and an image after noise removal; Figure 5 : The present invention provides a binary image of a method for filling a two-dimensional grain boundary defect based on three-dimensional information; Figure 6 : Schematic diagram of iterative corrosion of a method for filling two-dimensional grain boundary defects based on three-dimensional information in the present invention; Figure 7 : The invention discloses a grain center marking point image of a method for filling a two-dimensional grain boundary defect based on three-dimensional information; Figure 8 : The present invention provides a binary complete grain boundary image based on a method for filling two-dimensional grain boundary gaps based on three-dimensional information. DETAILED DESCRIPTION

[0029] The present invention provides a method for filling two-dimensional grain boundary gaps based on three-dimensional information. By combining three-dimensional information collection, missing area analysis and modeling, grain boundary filling algorithm, filling result verification, optimization and adjustment, etc., efficient and accurate grain boundary repair is achieved. This method not only improves the accuracy of grain boundary reconstruction and the reliability of material performance prediction, but also reduces experimental and computational costs, improves data processing efficiency, and has wide applicability. The present invention mainly involves the following aspects: 1. 3D information collection The three-dimensional structural information of crystalline materials is obtained by using high-resolution two-dimensional imaging technologies, such as optical microscopes and scanning electron microscopes. This information includes the morphology and position of grain boundaries and specific information about missing parts, providing basic information support for subsequent grain boundary filling.

[0030] 2. Missing area analysis and modeling Based on the collected 3D information, computer image processing techniques, such as image segmentation and feature extraction, are used to analyze the missing grain boundary areas. At the same time, 3D modeling techniques, such as geometric modeling and physical modeling, are used to model the missing areas and accurately identify the specific location and shape of the missing areas.

[0031] 3. Grain boundary filling algorithm A grain boundary filling algorithm based on three-dimensional information is designed, which can automatically or semi-automatically generate the grain boundary information of the missing part. The core of the algorithm is to deduce a reasonable filling scheme based on the morphological characteristics of the surrounding grain boundaries, such as grain boundary curvature, grain boundary orientation, etc., and crystal structure rules, such as crystal symmetry rules, crystal growth rules, etc. In this way, the consistency of the filled grain boundaries with the surrounding grain boundaries in morphology and structure can be ensured.

[0032] 4. Verification of filling results The filling results are verified by using 3D reconstruction technology, such as volume rendering technology and surface reconstruction technology. By comparing the filled grain boundaries with the overall structure of the crystal, it is ensured that the filled grain boundaries meet the overall structure and performance requirements of the crystal. If the verification results do not meet the requirements, the filling algorithm will be dynamically optimized and adjusted according to the verification feedback until the filling quality reaches the expected effect.

[0033] 5. Optimization and adjustment During the filling process, the algorithm is dynamically optimized according to the filling effect. The optimization content includes parameter adjustment optimization, algorithm structure optimization, etc. Through continuous optimization and adjustment, the accuracy and efficiency of the filling algorithm are improved to ensure the stability and reliability of the filling quality.

[0034] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.

[0035] like Figure 1 As shown, combined with Figure 2-Figure 8 The schematic diagram of the present invention shown is a method for filling two-dimensional grain boundary defects based on three-dimensional information, comprising: S1: Read in a series of metallographic image stacks of the sample (as shown in Figure 2, this operation can generate a series of metallographic image stacks to provide basic data for subsequent processing): A stack of 132 sample metallographic images with a size of 105×880×390 μm was read in, and all metallographic images were benchmarked and registered to correlate spatial information.

[0036] S1.1 After heat treatment, the sample is inlaid, ground, marked with diamond points, polished, etched, photographed, etc. to obtain the first metallographic image. The polishing, etching, photographing operations are repeated to obtain the second, third, and other series of metallographic images of the same area.

[0037] S1.2 In the process of processing specific areas, image j software is used to perform rigorous image processing operations on a series of grain metallographic images. Specifically, the diamond markers are used as references, and plugins, registration, and SIFT commands are used to ensure that all images are accurately aligned in the upper and lower positions to achieve consistency and accuracy in image analysis. A series of image stacks can be generated by accurately cropping the required area. (See Figure 2 ) S2: Hand-drawn complete metallographic diagram of the first layer: Preprocess all metallographic images.

[0038] S2.1 Manually improve the first metallographic image to ensure the integrity of its grain boundaries. The grain boundary filling work of all subsequent metallographic images can be completed automatically by computer; S2.2 grayscales all metallographic images, performs image enhancement and image smoothing filtering; S2.2.1 Grayscale processing uses floating point calculation method; S2.2.2 Image enhancement uses equalization technology formula; S2.2.3 Image smoothing filtering uses Gaussian filtering and wavelet transform.

[0039] S3: Remove noise and perform binarization before thinning (Figure 4 shows the image effect after noise removal. Figure 5 A binary image is presented. After these processing, the image is more conducive to subsequent analysis). S3.1 uses the imfill function to eliminate the noise in the metallographic image; S3.2 performs binarization operation on the metallographic image using the formula; S3.3 uses MATLAB to refine the undesirable phenomena such as inconsistent grain boundary width and burrs.

[0040] S4: Iteratively etch each grain in the metallographic image to find the grain center and obtain multiple seed record mark points (Figure 6 illustrates the iterative etching process, and Figure 7 is the final grain center mark point image. These results lay the foundation for further analysis of grain relationships). S4.1 sets different thresholds disk, and performs multi-scale iterative corrosion processing on the given image while maintaining the geometric features of the grain boundary; in this process, each time the corrosion operation is completed, the boundary of the grain will shrink one layer inward until the grain is in an iterative state before it completely disappears, and the center point of the grain is obtained; S4.2 uses the mathematical model Gaussian curvature or mean curvature to describe the grain boundary curvature, calculates the degree of curvature through the local grain boundary geometric characteristics, and infers the shape of the filled area and the surface structure.

[0041] S4.2.1 The curvature is calculated through the local grain boundary tangent plane equation and numerical methods, and the shape of the filled area is dynamically adjusted. The crystal symmetry and space group information are combined to ensure that the lattice structure of the filled area is consistent with the surrounding area.

[0042] S4.2.2 Use clustering algorithms to cluster crystal structure regions and analyze the similarities and differences between clusters. Combine clustering results with physical properties to build a data-driven similarity assessment framework. By adjusting the similarity threshold, accurately predict or fill in the characteristics of similar grains.

[0043] S4.2.3 For crystal structures, similarity metrics such as geometric distance, phase similarity, and energy minimization can be used to assign similarity scores to each grain to influence the filling strategy. Through a dynamic adjustment mechanism, the filling threshold and similarity criteria are adjusted in real time based on the grain distribution characteristics. Incremental learning is used to optimize parameter configuration as data accumulates and the algorithm iterates.

[0044] S4.3 records the center point of each grain before it disappears, and obtains multiple seed recording mark points.

[0045] S5: Track each grain in eight directions to calculate the adjacent grain ratio parameters. S5.1 implements a tracking strategy of eight positions (45° apart) for each marker point; S5.2 Ensure that the line is traced continuously from the marked point until the grain boundary is encountered, and the length of each line segment is recorded and the proportional parameter is calculated.

[0046] S6: Perform edge detection and extraction on the first layer of the complete metallographic image and then scale it according to the proportion; S6.1 performs edge detection and contour extraction on each grain in the first layer of the complete metallographic image; S6.2 is scaled according to the ratio parameters of two adjacent layers.

[0047] S7: filling each grain of the second layer metallographic image in turn to obtain a complete metallographic image of the second layer; S7.1 Use the program to fill a die; S7.2 calculates and selects the optimal Boundary IoU metric to evaluate the matching degree between the object boundary in the segmentation result and the true boundary and measure the boundary accuracy.

[0048] S7.3 loop fills all the grains of a layer of metallographic image.

[0049] S8: Obtain and output complete metallographic images of all layers in sequence. FIG8 is a binary complete grain boundary image finally output, which reflects the achievement of grain boundary filling achieved by the present invention.

[0050] The embodiments described above are merely descriptions of preferred implementation modes of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for filling two-dimensional grain boundary defects based on three-dimensional information, characterized in that: The following steps are involved: A series of metallographic images of the sample are collected using high-resolution two-dimensional imaging technology, and all metallographic images are benchmarked and registered based on specific markers using an image registration algorithm to accurately associate spatial information, wherein the image registration algorithm uses specific instructions of professional image processing software to achieve precise positioning; The grain boundaries of the first layer of metallographic images are hand-painted to make them complete, and then all metallographic images are preprocessed. The preprocessing operations include grayscale processing using floating-point calculation method, image enhancement based on equalization technology formula, and image smoothing filtering combined with Gaussian filtering and wavelet transform to ensure that the image quality meets the requirements of subsequent processing; Use specific functions to eliminate noise in metallographic images, perform binarization operations according to predetermined formulas, and then use professional software tools to refine problems such as inconsistent grain boundary width and burrs to obtain clear grain boundary images; By setting different threshold disks, multi-scale iterative corrosion operations are performed on each grain in the metallographic image while maintaining the geometric characteristics of the grain boundary. The center points of the grains are obtained at a specific iteration state before each grain completely disappears, and these center points are recorded as seed record markers to provide a basis for subsequent analysis. For each seed record mark point, eight azimuth tracking strategies with 45° intervals are implemented, starting from the mark point and continuously tracking until the grain boundary is encountered. In this process, the length of each line segment is accurately recorded, and then the ratio parameters of adjacent grains are calculated to determine the relative relationship between grains; Perform edge detection and contour extraction on each grain in the first layer of the complete metallographic image, and perform scaling operations based on the proportional parameters calculated between two adjacent layers to provide an accurate template for subsequent filling; A specially developed program is used to fill each grain of the second layer metallographic diagram in turn. During the filling process, intelligent judgment and operation are performed based on the results of previous calculations and processing, and the cycle is repeated until all grains of the layer are filled to obtain a complete metallographic diagram of the second layer. The above filling process is executed cyclically, and the metallographic images of subsequent layers are processed in turn until the complete metallographic images of all layers are output, so as to realize the comprehensive automatic filling of grain boundaries and ensure the accuracy and consistency of the filling results.

2. A method for filling two-dimensional grain boundary defects based on three-dimensional information, characterized in that: include: Step 1: read in a series of metallographic image stacks of the sample, and perform benchmark registration on all metallographic images to associate spatial information; Step 2: Hand-draw the first layer of complete grain boundary metallographic diagram and pre-process all metallographic diagrams; Step 3: Remove noise points and perform binarization before performing refinement processing; Step 4: Iteratively etch each grain in the metallographic image to find the grain center and obtain multiple seed record mark points; Step 5: Perform eight-direction tracking on each grain to calculate the proportional parameters of its adjacent layers; Step 6: Perform edge detection on each grain in the first layer of the complete metallographic image to extract the outline and scale it accordingly; Step 7: Fill each grain of the second layer metallographic image in turn to obtain a complete metallographic image of the second layer; Step 8: Obtain complete metallographic images of all layers in sequence and output them.

3. According to claim 2, a method for filling two-dimensional grain boundary defects based on three-dimensional information is characterized in that: The S1 includes: S1.1 After heat treatment, the sample is mounted, ground, marked with diamond points, polished, etched, photographed, etc. to obtain the first metallographic image. The polishing, etching, photographing operations are repeated to obtain the second, third, and other series of metallographic images of the same area; S1.2 In the process of processing specific areas, image j software is used to perform rigorous image processing operations on a series of metallographic images; specifically, diamond markers are used as references, and plugins, registration, and SIFT commands are used to ensure accurate alignment of all images in the upper and lower positions to achieve consistency and accuracy in image analysis; a series of image stacks are generated by accurately cropping the required areas.

4. According to claim 2, a method for filling two-dimensional grain boundary defects based on three-dimensional information is characterized in that: The S2 includes: S2.1 Manually improve the first metallographic image to ensure the integrity of its grain boundaries. The grain boundary filling work of all subsequent metallographic images can be completed automatically by computer; S2.2 grayscales all metallographic images, performs image enhancement and image smoothing filtering; S2.2.1 Grayscale processing uses floating point calculation method; S2.2.2 Image enhancement uses equalization technology formula; S2.2.3 Image smoothing filtering uses Gaussian filtering and wavelet transform.

5. The method for filling two-dimensional grain boundary defects based on three-dimensional information according to claim 2, characterized in that: The S3 includes: S3.1 uses the imfill function to eliminate the noise in the metallographic image; S3.2 performs binarization operation on the metallographic image using the formula; S3.3 uses MATLAB to refine the undesirable phenomena such as inconsistent grain boundary width and burrs.

6. The method for filling two-dimensional grain boundary defects based on three-dimensional information according to claim 2, characterized in that: The S4 includes: S4.1 sets different thresholds disk, and performs multi-scale iterative corrosion processing on the given image while maintaining the geometric features of the grain boundary; in this process, each time the corrosion operation is completed, the boundary of the grain will shrink one layer inward until the grain is in an iterative state before it completely disappears, and the center point of the grain is obtained; S4.2 uses the mathematical model Gaussian curvature or mean curvature to describe the grain boundary curvature, calculates the degree of curvature through the local grain boundary geometric characteristics, and infers the shape of the filled area and the surface structure; S4.2.1 Use the local grain boundary tangent plane equation and numerical method to calculate the curvature, dynamically adjust the shape of the filled area, and combine the crystal symmetry and space group information to ensure that the lattice structure of the filled area is consistent with the surrounding area; S4.2.2 Use clustering algorithms to cluster crystal structure regions and analyze the similarities and differences between clusters; combine clustering results with physical properties to build a data-driven similarity assessment framework; accurately predict or fill in the characteristics of similar grains by adjusting the similarity threshold; S4.2.3 For crystal structures, similarity metrics such as geometric distance, phase similarity, and energy minimization can be used to assign similarity scores to each grain to affect the filling strategy; through a dynamic adjustment mechanism, the filling threshold and similarity criteria can be adjusted in real time based on the grain distribution characteristics; incremental learning can be used to optimize parameter configuration as data accumulates and the algorithm iterates; S4.3 records the center point of each grain before it disappears, and obtains multiple seed recording mark points.

7. The method for filling two-dimensional grain boundary defects based on three-dimensional information according to claim 2, characterized in that: The S5 includes: S5.1 Implement a tracking strategy for each marker point in eight directions at 45° intervals; S5.2 Ensure that the line is traced continuously from the marked point until the grain boundary is encountered, and the length of each line segment is recorded and the proportional parameter is calculated.

8. The method for filling two-dimensional grain boundary defects based on three-dimensional information according to claim 2, characterized in that: The S6 includes: S6.1 performs edge detection and contour extraction on each grain in the first layer of the complete metallographic image; S6.2 is scaled according to the ratio parameters of two adjacent layers.

9. The method for filling two-dimensional grain boundary defects based on three-dimensional information according to claim 2, characterized in that: The S7 includes: S7.1 Use the program to fill a die; S7.2 calculates and selects the optimal Boundary IoU index to evaluate the matching degree between the object boundary in the segmentation result and the true boundary and measure the boundary accuracy; S7.3 loop fills all the grains of a layer of metallographic image.

10. The method for filling two-dimensional grain boundary defects based on three-dimensional information according to claim 2, characterized in that: After the complete metallographic images of all layers are obtained and output in sequence in S8, a step of verifying the filling results is also included. The verification method uses three-dimensional reconstruction technology to ensure that the filled grain boundaries meet the overall structure and performance requirements of the crystal. If the verification results do not meet the requirements, the filling algorithm is dynamically optimized and adjusted based on the verification feedback.

Images