Metal metallographic structure image classification display method and system

Abstract

The invention relates to the technical field of image display, in particular to a metal metallographic structure image classification display method and system. The method comprises the following steps: preprocessing a metal metallographic structure image, calculating and evaluating an image feature index Tzls, quantifying the overall quality of the metal metallographic structure image, automatically screening low-quality and repeated images, and reducing stored and processed invalid data; classifying the metal metallographic structure images by using two characteristics of structural factors and defect factors of the metal metallographic structure; the metal metallographic structure image is classified and graded by calculating and evaluating a structure characteristic index Jtzs and a defect characteristic index Qtzs; and all grade division results are transmitted to the image classification display module, so that timely and accurate data transmission is ensured, a user searches and filters according to a structure grade or a defect grade, a required image is quickly found, and the working efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of image display technology, specifically to a method and system for classifying and displaying metallographic images. Background Technology

[0002] Metallographic microstructure imaging devices are crucial equipment for observing and analyzing the internal microstructure of metallic materials. These devices allow for high-magnification observation of metal samples under a microscope, revealing information such as grain structure, phase composition, and defect distribution. This is of paramount importance for materials science research, quality control, and failure analysis. Traditional metallographic microscopes typically incorporate computer image processing technology, enabling high-definition image display, real-time analysis, and data storage, thus improving detection efficiency and accuracy. These devices are widely used in metallurgy, aerospace, and automotive manufacturing.

[0003] However, in practical applications, existing metallographic image display devices generate massive amounts of data from high-resolution images, requiring significant storage and processing power, which makes data management and analysis difficult. Furthermore, existing image processing and analysis algorithms are not accurate or efficient in automatically identifying and classifying complex microstructures, relying mainly on manual intervention, which increases the overall cost and the difficulty of using the equipment. At the same time, while metallographic image classification display devices have a wide range of applications, the application requirements of each field are different. For example, in the field of metal quality control, the first consideration is the defect characteristics; in the metallurgical industry, the first consideration is the structural characteristics of the metal. Existing metallographic image display devices cannot perform classification display, making them very inconvenient to use. Summary of the Invention

[0004] This invention proposes a method and system for classifying and displaying metallographic images, in order to solve the technical problems in the prior art, such as the large amount of high-resolution metallographic image data leading to difficulties in storage, processing, management and analysis; the low accuracy and efficiency of automatic identification and classification of complex microstructures, which relies on manual labor and thus increases the cost and difficulty of use; and the inability to classify and display structural features and defect features in a targeted manner according to the needs of different fields, resulting in inconvenience in use.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for classifying and displaying metallographic structures in metals includes the following steps: Batch image data acquisition of metallographic structures; Image processing algorithms are applied to preprocess image data, calculate and evaluate the image feature index Tzls, and filter out invalid data. Based on the image data after preprocessing and filtering out invalid data, the structural feature data of the metallographic structure is obtained, and the structural feature index Jtzs is calculated. The structural feature index Jtzs is evaluated, and the structural level of the metallographic structure is classified according to the evaluation results. Based on the image data after preprocessing and filtering out invalid data, the defect feature data in the image data is obtained and calculated to obtain the defect feature index Qtzs; the defect feature index Qtzs is evaluated, and the defect level of the metallographic structure is classified according to the evaluation results. Based on the structural grade classification and defect grade classification results of the metallographic structure, the images are displayed separately; users can search for and apply the images according to the level of structural grade or defect grade.

[0006] The batch image data acquisition of the metallographic structure specifically involves: image acquisition using an electron microscope; prior to image acquisition, focusing and exposure adjustment of the electron microscope; focusing adjustment to ensure the sample surface is centered in the image; exposure adjustment to ensure normal brightness and contrast; after electron microscope adjustment, the moving platform moves along a preset scanning path, irradiating the surface of the metallographic sample with an electron beam to form a scanning area, and scanning the metallographic structure sample; during this process, the electron microscope continuously acquires image data and saves it in real time through connected image acquisition software to obtain image data of the metallographic structure.

[0007] The image processing algorithm described above preprocesses the image data, specifically by removing random noise, adjusting brightness and darkness distribution, and emphasizing edges and details. Then, features are extracted from the image data to filter out invalid data. First, image feature data is collected, including image sharpness (Tqx), image resolution (Tfb), and image repetition (Tcf). Image sharpness (Tqx) is obtained through edge detection; image resolution (Tfb) is obtained using image processing software; and image repetition (Tcf) is obtained using a hash algorithm. After dimensionless processing of image sharpness (Tqx), image resolution (Tfb), and image repetition (Tcf), the image feature dataset is integrated.

[0008] Based on the image feature dataset, the image feature index Tzls is calculated. The formula for calculating the image feature index Tzls is as follows:

[0009] The image feature index Tzls is compared with the preset feature threshold Q to generate the following evaluation scheme: If the preset feature threshold Q is greater than or equal to the image feature index Tzls, it means that the current image quality is abnormal and there are duplicate images, which cannot be used for detection. In this case, the image is removed. If the preset feature threshold Q is less than the image feature index Tzls, it means that the current image quality is normal and there are no duplicate images. This image will be retained for subsequent image classification.

[0010] The acquisition of structural feature data of metallographic structures specifically involves: acquiring grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq from image data after preprocessing and filtering out invalid data using Python and OpenCV; and constructing a structural feature dataset of metallographic structures by performing dimensionless processing on grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq.

[0011] Based on the structural feature dataset of metallographic structures, the structural feature index Jtzs is calculated. The formula for calculating the structural feature index Jtzs is as follows:

[0012] The structural characteristic index Jtzs is compared with the first preset structural threshold W. 1 Second preset structural threshold W 2 The comparison yields the following evaluation results, where the first preset structural threshold W... 1 >Second preset structural threshold W 2 ; If the structural characteristic index Jtzs ≥ the first preset structural threshold W 1 If the percentage of a specific phase in the current metallographic image is 46%–65%, the uniform distribution is deemed unacceptable. In this case, the first evaluation result is generated, and the image is classified as a third-level structural image. If the first preset structural threshold W 1 > Structural characteristic index Jtzs > Second preset structural threshold W 2 If the value is 25%–45%, it means that the specific phase in the current metallographic image accounts for 25%–45%. At this point, a second evaluation result is generated, and the image is classified as a second-level structural image. If the second preset structural threshold W 2 If the structural characteristic index Jtzs is greater than or equal to 0.1%, it means that the proportion of the specific phase in the current metallographic image is 0.1%–24%. At this point, a third evaluation result is generated, and the image is classified as a first-level structural image.

[0013] The acquisition of defect feature data from image data specifically involves: constructing a defect feature dataset, which is constructed by dimensionlessly processing point defect index Dqx, line defect index Xqx, surface defect index Mqx, and micro defect index Wgq; the point defect index Dqx is obtained using image processing algorithms; the line defect index Xqx is obtained through edge detection; the surface defect index Mqx is obtained through segmentation algorithms; and the micro defect index Wgq is obtained through microscopic imaging technology.

[0014] Based on the defect feature dataset, the defect feature index Qtzs is calculated. The formula for calculating the defect feature index Qtzs is as follows:

[0015] In the formula, a1, a2, a3 and a4 represent the weight values ​​of the point defect index Dqx, the line defect index Xqx, the surface defect index Mqx and the micro defect index Wgq, respectively, and Zqz represents the total weight value; and 0 < a1, 0 < a2, 0 < a3, 0 < a4; The defect characteristic index Qtzs is compared with the first preset defect threshold E. 1 Second preset defect threshold E 2 For comparison, the first defect threshold E 1 >Second defect threshold E 2 The following evaluation results were generated: If the defect characteristic index Qtzs ≥ the first preset defect threshold E 1 If the image shows a large number of large defects in the current metallographic structure image, it indicates that the defects have a significant impact on the material properties. At this point, a fourth evaluation result is generated, and the image is classified as a third-level defect image. If the first preset defect threshold E 1 >Defect characteristic index Qtzs >Second preset defect threshold E 2 This indicates that in the current metallographic image, the number and size of defects are fewer and smaller compared to the third-level defect image, which has a certain impact on material properties; at this time, the fifth evaluation result is generated, and the image is classified as a second-level defect image. If the second preset defect threshold E 2 If the defect feature index Qtzs is greater than or equal to the defect size index Qtzs, it means that in the current metallographic image, the number and size of defects are less than those in the second-level defect image, and the defects have no effect on the material properties or their effect is negligible. At this time, the sixth evaluation result is generated, and the image is classified as a first-level defect image.

[0016] The classification results based on the structural and defect levels of the metallographic structure are displayed separately. Specifically, based on first-level, second-level, and third-level structural images, different structural level classification results are displayed, including a tree-structure diagram showing the classification results for different structural levels. Thumbnails of representative images are provided for each structural level, allowing users to click on the thumbnails to view larger images and detailed information. Based on first-level, second-level, and third-level defect images, defect level classification results are displayed, including a table listing different types of defects and their corresponding images. Users can search using structural or defect level as filters on the interactive interface.

[0017] A metallographic structure image classification and display system includes a metal image acquisition module, an automated analysis module, a first image classification module, a second image classification module, and an image classification and display module; The metal image acquisition module is used to acquire batch image data of metallographic structures. The automated analysis module is used to preprocess image data using image processing algorithms, calculate and evaluate the image feature index Tzls, and filter out invalid data. The first image classification module is used to obtain structural feature data of metallographic structure based on image data after preprocessing and filtering out invalid data, calculate structural feature index Jtzs, evaluate the structural feature index Jtzs, and classify the metallographic structure according to the evaluation results. The second image classification module is used to obtain defect feature data in the image data based on the preprocessed image data and after filtering out invalid data, and to calculate and obtain the defect feature index Qtzs; to evaluate the defect feature index Qtzs, and to classify the defect level of the metallographic structure according to the evaluation results. The image classification and display module is used to classify and display images according to the structural grade classification results and defect grade classification results of metallographic structures; users can search for and apply images according to the level of structural grade or defect grade.

[0018] Compared with the prior art, the technical solution of the present invention has the following beneficial effects: This invention provides a method for classifying and displaying metallographic images. It involves batch image acquisition of metallographic structures using an electron microscope. Before acquisition, targeted focusing and exposure adjustments are performed. Focusing ensures the sample surface is precisely centered in the image, while exposure adjustment maintains appropriate brightness and contrast. A moving platform then completes the scan along a preset scanning path, and image acquisition software saves the data in real time. This acquisition method ensures that the acquired images maintain uniformity in position, brightness, and contrast, avoiding image quality issues caused by improper acquisition parameters. The acquired batch image data possesses consistency and integrity, providing a stable and reliable data foundation for subsequent preprocessing, feature extraction, and classification, making subsequent processing more targeted and effective. When preprocessing the image data using image processing algorithms, operations such as eliminating random noise, adjusting brightness and darkness distribution, and emphasizing edges and details optimize the visual effect and data quality. Subsequently, image sharpness, resolution, and repetition feature data are extracted, and dimensionless processing is used to construct an image feature dataset. Then, an image feature index is calculated based on this dataset and compared with a preset threshold. This process accurately identifies and removes blurry, low-resolution, and duplicate invalid images, reducing the pressure on subsequent data storage, preventing invalid data from consuming storage resources, and eliminating interference from low-quality data on classification results. This ensures that all images used for classification are high-quality and valid data, laying the foundation for accurate extraction of subsequent structural and defect features and improving the efficiency of the overall data processing workflow. Preprocessing the metallographic images improves the accuracy of subsequent analysis and facilitates more precise feature extraction. Secondly, calculating the image feature index Tzls quantifies the overall quality of the metallographic images. Finally, by comparing a preset feature threshold Q with Tzls, low-quality and duplicate images are automatically filtered out, retaining high-quality, non-duplicated images. This ensures the accuracy and reliability of subsequent image classification and analysis, reduces the storage and processing of invalid data, and improves resource utilization efficiency.

[0019] Furthermore, metallographic images are classified using two main characteristics: structural factors and defect factors. First, based on preprocessed effective image data, the structural features of the metallographic images are quantified by acquiring and calculating the structural feature index Jtzs. Second, a preset structural threshold W is used... 1 and structural threshold W 2The system compares the results with the structural characteristic index Jtzs to generate corresponding evaluation results. Based on these results, it accurately classifies the metallographic structure into different levels for easy classification and display. Finally, it quantifies the defect factors in the metallographic images by calculating the defect characteristic index Qtzs to accurately classify the defect levels. All classification results are then transmitted to the image classification and display module, ensuring timely and accurate data transmission. Users can search and filter by structural or defect level to quickly find the desired images, improving work efficiency.

[0020] Furthermore, this invention categorizes and displays the structural level classification results and defect level classification results in different formats. The structural level is presented using a tree structure diagram paired with representative image thumbnails, while the defect level lists different defect types and their corresponding images in a table format. It also provides users with search functions based on either structural or defect level. The combination of the tree structure diagram and thumbnails allows users to quickly browse the overall classification of different structural levels. Clicking on a thumbnail allows users to view a larger image and detailed information. The table format facilitates intuitive viewing of images corresponding to various defects, while the level search function allows users to directly filter images that meet their needs. This classification display and search method precisely adapts to the personalized needs of different application fields. Users do not need to sift through massive amounts of images one by one; they can quickly locate the target image, significantly improving the convenience and efficiency of image retrieval and reducing the operational difficulty of the device. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the block structure of a metallographic structure image classification and display system according to the present invention. Figure 2 This is a schematic diagram of the process structure of a metallographic structure image classification and display method according to the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example 1 Please see Figure 2 This embodiment provides a method for classifying and displaying metallographic images, including the following steps: Batch image data acquisition of metallographic structures using electron microscopy ensures that the acquired images are clear and complete, providing a high-quality data foundation for subsequent classification and analysis; Image processing algorithms are applied to preprocess image data, calculate and evaluate the image feature index Tzls, filter out invalid data, reduce the interference of invalid data on subsequent classification results, and improve the accuracy of classification analysis. Based on the image data after preprocessing and filtering out invalid data, the structural feature data of metallographic structures are obtained, and the structural feature index Jtzs is calculated. The structural feature index Jtzs is evaluated, and the structural level of metallographic structures is classified according to the evaluation results. The structural differences of different metallographic structures are clarified, providing a basis for subsequent targeted applications. Based on the image data after preprocessing and filtering out invalid data, defect feature data in the image data is obtained and calculated to obtain the defect feature index Qtzs; the defect feature index Qtzs is evaluated, and the defect level of the metallographic structure is classified according to the evaluation results, covering the dual core dimensions of structure and defects, making the classification more complete. Based on the structural and defect classification results of the metallographic structure, images are displayed in categories. Users can search for and apply images through the user interface according to the structural or defect level, thus improving the efficiency of image retrieval and practical application.

[0024] In this embodiment, images of metallographic samples are acquired using an electron microscope; the acquired image data is preprocessed and its quality is assessed; the samples are classified based on the structural characteristics of the metallographic structure; the samples are also classified based on the defect characteristics of the metallographic structure; finally, the classification results are displayed, and a user interface is provided.

[0025] Example 2 Based on the metallographic structure image classification and display method proposed in Embodiment 1, this embodiment proposes a metallographic structure image classification and display system to implement the method steps of Embodiment 1, such as... Figure 1 As shown, it includes a metal image acquisition module, an automated analysis module, a first image classification module, a second image classification module, and an image classification display module; The metal image acquisition module is used to acquire batch image data of metal metallographic structures using an electron microscope. Before acquiring the images, the electron microscope is focused and the exposure is adjusted, the moving platform is used to scan the metal sample, and finally the image acquisition software is used to acquire images in batches. The automated analysis module is used to preprocess and extract the acquired images; it applies image processing algorithms to preprocess the images, including noise reduction, contrast enhancement, and sharpening; and then calculates and evaluates the image feature index Tzls to filter out low-quality and duplicate images. The first image classification module is used to obtain structural feature data of metallographic structure based on image data after preprocessing and filtering out invalid data, calculate structural feature index Jtzs, evaluate the structural feature index Jtzs, and classify the metallographic structure according to the evaluation results. The second image classification module is used to obtain defect feature data in the image data based on the preprocessed image data and after filtering out invalid data, and to calculate and obtain the defect feature index Qtzs; to evaluate the defect feature index Qtzs, and to classify the defect level of the metallographic structure according to the evaluation results. The image classification and display module is used to display the classification results and provide a user interface; it classifies and displays the results according to the structural grade classification results and defect grade classification results of the metallographic structure; users can search for and apply images according to the level of structural grade or defect grade.

[0026] In this embodiment, the metal image acquisition module is used to scan metal samples using an electron microscope to acquire images; the automated analysis module is used to quantify image quality and filter the acquired images, removing low-quality and duplicate images; the first image classification module is used to quantify the structural features of the metallographic structure, calculating and evaluating the structural feature index Jtzs to classify the structure level; the second image classification module is used to quantify the defect features of the metallographic structure image, calculating and evaluating the defect feature index Qtzs to classify the defect level; the classification display module is used to display the classification results and provide a user interface, allowing users to search for and apply images based on the classification results of structural and defect levels.

[0027] In a preferred embodiment of the present invention, the metal image acquisition module is used to acquire metal images based on an electron microscope; by adjusting the focus of the electron microscope, the sample surface is displayed in the center of the image; and by adjusting the exposure, the brightness and contrast of the image are normalized; after the electron microscope is adjusted, the moving platform moves according to a preset scanning path to irradiate the metal sample surface with an electron beam to form a scanning area; during this process, the electron microscope continuously acquires image data and saves it in real time through the connected image acquisition software.

[0028] In this embodiment, firstly, the focusing knob of the microscope is adjusted so that the sample surface is clearly imaged in the center of the electron microscope's field of view; secondly, the exposure time and light source intensity of the microscope are adjusted so that the image brightness is moderate and there is no overexposure or underexposure; during scanning on the moving platform, the scanning path is preset according to the size of the sample and the research needs to ensure full coverage of the sample; finally, the electron microscope transmits the acquired metallographic structure image data to the image acquisition software through the data interface.

[0029] In a preferred embodiment of the present invention, the automated analysis module includes a preprocessing unit, an image feature acquisition unit, a first calculation unit, and a first evaluation unit; the preprocessing unit is used to improve the visual clarity and observation effect of the image; the specific steps of the preprocessing unit include eliminating random noise in the image, adjusting the brightness and darkness distribution of the image, and emphasizing the edges and details in the image; The image feature acquisition unit acquires the following data: image sharpness Tqx, image resolution Tfb, and image repetition Tcf. Image sharpness Tqx is obtained through edge detection; image resolution Tfb is obtained through image processing software; and image repetition Tcf is obtained through a hash algorithm. After dimensionless processing of image sharpness Tqx, image resolution Tfb, and image repetition Tcf, the data are integrated into an image feature dataset.

[0030] The first calculation unit extracts image sharpness Tqx, image resolution Tfb, and image repetition Tcf, and calculates the image feature index Tzls using the following formula:

[0031] The first evaluation unit generates the following evaluation scheme by comparing a preset feature threshold Q with the image feature index Tzls: If the preset feature threshold Q is greater than or equal to the image feature index Tzls, it means that the current image quality is abnormal and there are duplicate images, which cannot be used for detection. In this case, the image is removed. If the preset feature threshold Q is less than the image feature index Tzls, it means that the current image quality is normal and there are no duplicate images. This image will be retained for subsequent image classification.

[0032] In this embodiment, the preprocessing unit eliminates random noise in the image, adjusts the brightness and darkness distribution of the image, and emphasizes the edges and details in the image. The purpose is to smooth the image, improve the overall image clarity, and enhance the local features of the image. The image sharpness Tqx is acquired to make the microstructure and features of the metallographic structure more clearly visible, facilitating observation and analysis. The image resolution Tfb is acquired because high-resolution images contain more detailed information and can more accurately reflect the microstructure of the metallographic structure. The image repeatability Tcf is acquired to avoid duplicate images in the dataset, thereby improving data diversity and the effectiveness of analysis.

[0033] In a preferred embodiment of the present invention, the first image classification module includes a structural feature acquisition unit, a second calculation unit, a second evaluation unit, and a structural level transmission unit; the structural feature acquisition unit acquires grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq using Python and OpenCV; and simultaneously performs dimensionless processing on the grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq to construct a structural feature dataset.

[0034] The second calculation unit extracts the grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq, and calculates the structural feature index Jtzs using the following formula:

[0035] The second evaluation unit uses a preset first preset structural threshold W. 1 Second preset structural threshold W 2 The structure is compared with the structural feature index Jtzs, and the first preset structural threshold W is used. 1 >Second preset structural threshold W 2 The following evaluation results were generated; If the structural characteristic index Jtzs ≥ the first preset structural threshold W 1 If the structure-specific phase accounts for 46%–65% of the current metallographic image, it is unevenly distributed. At this point, the first evaluation result is generated, and the image is classified as a third-level structure image. The structure-specific phase is represented by the region in the metallographic structure that contains grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq. If the first preset structural threshold W 1 > Structural characteristic index Jtzs > Second preset structural threshold W 2 This indicates that in the current metallographic image, the specific phase of the structure accounts for 25%–45%, and the distribution is more uniform compared to the third-level structure image; at this time, a second evaluation result is generated, and the image is classified as a second-level structure image; If the second preset structural threshold W 2 If the structural characteristic index Jtzs is greater than or equal to the current metallographic image, it means that the proportion of the structural specific phase in the image is 0.1%–24%, and the distribution is more uniform compared to the second-level structural image. At this point, the third evaluation result is generated, and the image is classified as a first-level structural image. The structure level transmission unit transmits the division results of the first-level structure image, the second-level structure image, and the third-level structure image to the image classification display module, respectively.

[0036] The grain area Jmj is collected to evaluate the effect of heat treatment processes, such as annealing and quenching, on grain size; the grain boundary density Jmd is collected to evaluate the corrosion resistance and fatigue resistance of the material, because grain boundaries are used to prevent corrosion and crack propagation; the pore area Kxm represents the defects in the metallic material, and their size and number directly affect the strength and fatigue life of the material; the texture intensity Wlq reflects the arrangement of grains in a specific direction, which has an important impact on the mechanical and physical properties of the material.

[0037] In this embodiment, the above-described evaluation method can effectively classify metallographic images based on the structural feature index Jtzs and the first preset structural threshold W. 1 Second preset structural threshold W 2 By comparing different phases, the proportion and distribution of specific phases in the structure are determined, thereby generating corresponding evaluation results and structural image levels. This method helps to more accurately analyze the characteristics of metallic metallographic structures, providing a basis for material performance evaluation and process optimization.

[0038] In a preferred embodiment of the present invention, the second image classification module includes a defect feature acquisition unit, a third calculation unit, a third evaluation unit, and a defect level transmission unit; the defect feature acquisition unit is used to extract defect feature datasets from metallographic images; the defect feature dataset is constructed by dimensionless processing of point defect index Dqx, line defect index Xqx, surface defect index Mqx, and micro defect index Wgq; the point defect index Dqx is obtained by using image processing algorithms; the line defect index Xqx is obtained by edge detection; the surface defect index Mqx is obtained by segmentation algorithms; and the micro defect index Wgq is obtained by microscopic imaging technology.

[0039] Based on the third calculation unit, the defect feature index Qtzs is calculated by extracting the point defect index Dqx, line defect index Xqx, surface defect index Mqx, and micro defect index Wgq using the following formula:

[0040] In the formula, a1, a2, a3 and a4 represent the weight values ​​of the point defect index Dqx, the line defect index Xqx, the surface defect index Mqx and the micro defect index Wgq, respectively, and Zqz represents the total weight value; and 0 < a1, 0 < a2, 0 < a3, 0 < a4; The third evaluation unit uses a first preset defect threshold E 1 Second preset defect threshold E 2 Compared with the defect feature index Qtzs, where the first preset defect threshold E 1 >Second preset defect threshold E 2 The following evaluation results were generated: If the preset defect characteristic index Qtzs ≥ the first preset defect threshold E 1 If the image shows a large number of large defects in the current metallographic structure image, it indicates that the defects have a significant impact on the material properties. At this point, a fourth evaluation result is generated, and the image is classified as a third-level defect image. If the first preset defect threshold E 1 >Defect characteristic index Qtzs >Second preset defect threshold E 2 This indicates that in the current metallographic image, the number and size of defects are fewer and smaller compared to the third-level defect image, which has a certain impact on material properties; at this time, the fifth evaluation result is generated, and the image is classified as a second-level defect image. If the second preset defect threshold E 2 If the defect feature index Qtzs is greater than or equal to the defect feature index Qtzs, it means that in the current metallographic image, the number and size of defects are less and smaller than those in the second-level defect image, and have no or negligible impact on the material properties; at this time, the sixth evaluation result is generated and the image is classified as a first-level defect image. The defect level transmission unit transmits the classification results of the first-level defect image, the second-level defect image, and the third-level defect image to the image classification display module, respectively.

[0041] In this embodiment, the point defect index Dqx refers to the number and distribution of tiny point defects in a metallic material. A large number of point defects can reduce the strength and toughness of the material, especially under high stress or fatigue conditions. The line defect index Xqx refers to the number and characteristics of line defects in a metallic material. Line defects are prone to becoming crack initiation points, affecting the durability and lifespan of the material during use. The surface defect index Mqx refers to the number and size of large-area defects on or inside a metallic material. The presence of surface defects requires appropriate processing and treatment measures to ensure that the quality of the final product meets the requirements. The microscopic defect index Wgq refers to the defects in the microstructure inside a metallic material, including abnormal grain growth and phase transformation regions. The appearance of defects can lead to the non-uniformity and instability of the material, affecting its behavior under different temperatures and stresses.

[0042] In a preferred embodiment of the present invention, the image classification display module includes a structure level display unit, a defect level display unit, and a user interface; the structure level display unit displays different structure level classification results based on a first-level structure image, a second-level structure image, and a third-level structure image, including using a tree structure diagram to display the classification results of different structure levels, and setting a thumbnail of a representative image under each structure level for users to click on the thumbnail to view a larger image and detailed information; The defect level display unit displays the defect level classification results based on the first-level defect image, the second-level defect image, and the third-level defect image, including listing different types of defects and their corresponding images in a table format; The user interface is used to search and view metallographic images; users can search based on structural grade or defect grade as filtering conditions.

[0043] In this embodiment, through the structural level display unit and the defect level display unit, users can select according to the characteristics of different metallographic structures, quickly understand the characteristics and differences of different features and levels of metallographic structures, help users accurately evaluate metal materials, and save users time and effort. Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for classifying and displaying metallographic images, characterized in that: Includes the following steps: Batch image data acquisition of metallographic structures; Image processing algorithms are applied to preprocess image data, calculate and evaluate the image feature index Tzls, and filter out invalid data. Based on the image data after preprocessing and filtering out invalid data, the structural feature data of the metallographic structure is obtained, and the structural feature index Jtzs is calculated. The structural feature index Jtzs is evaluated, and the structural level of the metallographic structure is classified according to the evaluation results. Based on the image data after preprocessing and filtering out invalid data, the defect feature data in the image data is obtained and calculated to obtain the defect feature index Qtzs. The defect characteristic index Qtzs is evaluated, and the defect level is classified according to the evaluation results of the metallographic structure. Based on the structural grade classification results and defect grade classification results of the metallographic structure, they are classified and displayed separately; Users search for and apply images based on their structural or defect level.

2. The method for classifying and displaying metallographic images according to claim 1, characterized in that, The batch image data acquisition of the metallographic structure specifically involves: image acquisition using an electron microscope; prior to image acquisition, focusing the electron microscope to ensure the sample surface is centered in the image; and then adjusting the exposure to ensure normal brightness and contrast. After the electron microscope is adjusted, the moving platform moves according to the preset scanning path, irradiating the surface of the metallographic sample with an electron beam to form a scanning area and scanning the metallographic structure sample. During this process, the electron microscope continuously acquires image data and saves it in real time through the connected image acquisition software to obtain image data of the metallographic structure.

3. The method for classifying and displaying metallographic images according to claim 1, characterized in that, The image processing algorithm described above preprocesses the image data, specifically by removing random noise, adjusting brightness and darkness distribution, and emphasizing edges and details. Then, features are extracted from the image data to filter out invalid data. First, image feature data is collected, including image sharpness (Tqx), image resolution (Tfb), and image repetition (Tcf). Image sharpness (Tqx) is obtained through edge detection; image resolution (Tfb) is obtained using image processing software; and image repetition (Tcf) is obtained using a hash algorithm. After dimensionless processing of image sharpness (Tqx), image resolution (Tfb), and image repetition (Tcf), the image feature dataset is integrated.

4. The method for classifying and displaying metallographic structures according to claim 3, characterized in that, Based on the image feature dataset, the image feature index Tzls is calculated. The formula for calculating the image feature index Tzls is as follows: The image feature index Tzls is compared with the preset feature threshold Q to generate the following evaluation scheme: If the preset feature threshold Q is greater than or equal to the image feature index Tzls, it means that the current image quality is abnormal and there are duplicate images, which cannot be used for detection. In this case, the image is removed. If the preset feature threshold Q is less than the image feature index Tzls, it means that the current image quality is normal and there are no duplicate images. This image will be retained for subsequent image classification.

5. The method for classifying and displaying metallographic images according to claim 1, characterized in that, The acquisition of structural feature data of metallographic structures specifically involves: acquiring grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq from image data after preprocessing and filtering out invalid data using Python and OpenCV; and constructing a structural feature dataset of metallographic structures by performing dimensionless processing on grain area Jmj, grain boundary density Jmd, pore area Kxm, and texture intensity Wlq.

6. The method for classifying and displaying metallographic images according to claim 5, characterized in that, Based on the structural feature dataset of metallographic structures, the structural feature index Jtzs is calculated. The formula for calculating the structural feature index Jtzs is as follows: The structural characteristic index Jtzs is compared with the first preset structural threshold W. 1 Second preset structural threshold W 2 The comparison yields the following evaluation results, where the first preset structural threshold W... 1 >Second preset structural threshold W 2 ; If the structural characteristic index Jtzs ≥ the first preset structural threshold W 1 If the percentage of a specific phase in the current metallographic image is 46%–65%, the uniform distribution is deemed unacceptable. In this case, the first evaluation result is generated, and the image is classified as a third-level structural image. If the first preset structural threshold W 1 > Structural characteristic index Jtzs > Second preset structural threshold W 2 If the value is 25%–45%, it means that the specific phase in the current metallographic image accounts for 25%–45%. At this point, a second evaluation result is generated, and the image is classified as a second-level structural image. If the second preset structural threshold W 2 If the structural characteristic index Jtzs is greater than or equal to 0.1%, it means that the proportion of the specific phase in the current metallographic image is 0.1%–24%. At this point, a third evaluation result is generated, and the image is classified as a first-level structural image.

7. The method for classifying and displaying metallographic images according to claim 1, characterized in that, The acquisition of defect feature data from image data specifically involves: constructing a defect feature dataset, which is constructed by dimensionlessly processing point defect index Dqx, line defect index Xqx, surface defect index Mqx, and micro defect index Wgq; the point defect index Dqx is obtained using image processing algorithms; the line defect index Xqx is obtained through edge detection; the surface defect index Mqx is obtained through segmentation algorithms; and the micro defect index Wgq is obtained through microscopic imaging technology.

8. The method for classifying and displaying metallographic images according to claim 7, characterized in that, Based on the defect feature dataset, the defect feature index Qtzs is calculated. The formula for calculating the defect feature index Qtzs is as follows: In the formula, a1, a2, a3 and a4 represent the weight values ​​of the point defect index Dqx, the line defect index Xqx, the surface defect index Mqx and the micro defect index Wgq, respectively, and Zqz represents the total weight value; and 0 < a1, 0 < a2, 0 < a3, 0 < a4; The defect characteristic index Qtzs is compared with the first preset defect threshold E. 1 Second preset defect threshold E 2 For comparison, the first defect threshold E 1 >Second defect threshold E 2 The following evaluation results were generated: If the defect characteristic index Qtzs ≥ the first preset defect threshold E 1 If the image shows a large number of large defects in the current metallographic structure image, it indicates that the defects have a significant impact on the material properties. At this point, a fourth evaluation result is generated, and the image is classified as a third-level defect image. If the first preset defect threshold E 1 >Defect characteristic index Qtzs >Second preset defect threshold E 2 This indicates that in the current metallographic image, the number and size of defects are fewer and smaller compared to the third-level defect image, which has a certain impact on material properties; at this time, the fifth evaluation result is generated, and the image is classified as a second-level defect image. If the second preset defect threshold E 2 If the defect feature index Qtzs is greater than or equal to the defect size index Qtzs, it means that in the current metallographic image, the number and size of defects are less than those in the second-level defect image, and the defects have no effect on the material properties or their effect is negligible. At this time, the sixth evaluation result is generated, and the image is classified as a first-level defect image.

9. The method for classifying and displaying metallographic images according to claim 1, characterized in that, The classification results based on the structural and defect levels of the metallographic structure are displayed separately. Specifically, based on first-level, second-level, and third-level structural images, different structural level classification results are displayed, including a tree-structure diagram showing the classification results for different structural levels. Thumbnails of representative images are provided for each structural level, allowing users to click on the thumbnails to view larger images and detailed information. Based on first-level, second-level, and third-level defect images, defect level classification results are displayed, including a table listing different types of defects and their corresponding images. Users can search using structural or defect level as filters on the interactive interface.

10. A metallographic structure image classification and display system, based on the metallographic structure image classification and display method according to any one of claims 1 to 9, characterized in that, It includes a metal image acquisition module, an automated analysis module, a first image classification module, a second image classification module, and an image classification display module; The metal image acquisition module is used to acquire batch image data of metallographic structures. The automated analysis module is used to preprocess image data using image processing algorithms, calculate and evaluate the image feature index Tzls, and filter out invalid data. The first image classification module is used to obtain structural feature data of metallographic structure based on image data after preprocessing and filtering out invalid data, calculate structural feature index Jtzs, evaluate the structural feature index Jtzs, and classify the metallographic structure according to the evaluation results. The second image classification module is used to obtain defect feature data in the image data based on the preprocessed image data and after filtering out invalid data, and to calculate the defect feature index Qtzs. The defect characteristic index Qtzs is evaluated, and the defect level is classified according to the evaluation results of the metallographic structure. The image classification and display module is used to classify and display the results based on the structural level classification results and defect level classification results of the metallographic structure. Users search for and apply images based on their structural or defect level.

Images