A method and system for evaluating the tensile resistance of an arc contact image

Abstract

The application discloses a kind of arc contact image tensile resistance evaluation processing method and system, the method includes the following operating steps: the tensile load image sequence of arc contact is collected, the tensile load image sequence of arc contact is preprocessed, and the image sequence of arc contact is obtained to be detected arc contact image;Arc contact surface is sprayed with uniform optical speckle;The stress and strain field of the tensile load of arc contact is obtained to the arc contact image to be detected in image sequence;Through stress and strain field, the fatigue degree of the arc contact to be detected is judged;Through fatigue degree, the tensile resistance of the arc contact to be detected is obtained;Stress and strain field include speckle strain field and stable strain field;Based on the tensile resistance of the arc contact to be detected, the evaluation report of arc contact is generated.

Description

Technical Field

[0001] This invention relates to the field of tensile strength testing of arc contacts, and more particularly to a method and system for evaluating the tensile strength of arc contact images. Background Technology

[0002] Arc contacts, as key components in electrical equipment, are widely used in power systems, such as circuit breakers and switchgear. Their primary function is to contact the circuit and bear the load under high current conditions, while simultaneously providing rapid interruption when an electric arc is generated, ensuring the normal operation of the power system. The performance of arc contacts directly affects the reliability, safety, and service life of the equipment; therefore, assessing their tensile strength and fatigue is crucial for extending equipment life, reducing maintenance costs, and preventing failures.

[0003] Traditionally, the performance evaluation of arc contacts has relied primarily on manual inspection and periodic mechanical testing. However, existing inspection methods often have several shortcomings. First, manual inspection cannot provide sufficient accuracy and consistency, and it is difficult to effectively identify micro-cracks or fatigue damage. Second, traditional mechanical testing methods often rely on destructive testing, which cannot capture the deformation process of the arc contact under dynamic loads in real time. Therefore, existing methods struggle to achieve real-time monitoring and prediction of the long-term health status of arc contacts, and it is difficult to detect potential fatigue damage or cracks in advance.

[0004] In recent years, with the development of computer vision and image processing technologies, image analysis-based non-destructive testing methods have been gradually applied to the inspection of arc contacts. These methods capture the deformation process of arc contacts under external forces in real time using high-speed cameras and combine image processing techniques to extract stress and strain fields, thereby analyzing the fatigue of arc contacts. However, in these existing technologies, how to efficiently and accurately obtain the stress and strain fields of arc contacts and combine them with fatigue assessment to evaluate their tensile strength remains an urgent problem to be solved.

[0005] Currently, research on the tensile strength assessment of arc contacts still faces several challenges. First, accurately predicting the degree of fatigue damage to arc contacts from strain fields in image sequences, especially their deformation behavior under high-frequency dynamic loads, requires high-precision image acquisition and analysis techniques. Second, accurately assessing the tensile strength of arc contacts and predicting their remaining service life based on stress and strain field analysis results remains a complex engineering problem. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for evaluating the tensile strength of arc contact images, which solves the aforementioned technical problems pointed out in the prior art.

[0007] This invention provides a method for evaluating the tensile strength of an arc contact image, comprising the following steps:

[0008] Acquire a tensile load image sequence of the arc contact, preprocess the tensile load image sequence of the arc contact, and obtain the image sequence of the arc contact to be detected;

[0009] The surface of the arc contact is coated with a uniform optical speckle pattern;

[0010] The stress and strain fields of the arc contact under tensile load are obtained from the images of the arc contact to be detected in the image sequence; the fatigue degree of the arc contact to be detected is determined by the stress and strain fields; and the tensile strength of the arc contact to be detected is obtained by the fatigue degree.

[0011] The stress and strain fields include speckle strain fields and steady strain fields;

[0012] An evaluation report for the arc contact is generated based on the tensile strength of the arc contact to be tested.

[0013] Preferably, the step of obtaining the stress and strain field of the tensile load of the arc contact in the image of the arc contact to be detected includes:

[0014] For each image of the arc contact to be detected in the image sequence, the pixels with the same gray value are clustered to obtain pixel clusters; the dispersion of the pixel clusters is analyzed, and the speckle region and stable region in the image of the arc contact to be detected are obtained based on the dispersion of the pixel clusters; displacement field operation is performed on the speckle region and stable region to obtain speckle strain field and stable strain field.

[0015] Preferably, the fatigue degree of the contact under test is calculated using the stress and strain field. The specific operation steps are as follows:

[0016] The strain field deviation value is calculated between the stable strain field and the speckle strain field. The strain field deviation value is used to extract the strain field abrupt change region of the arc contact image to be tested. The crack profile is extracted through the strain field abrupt change region. The crack distribution density of the crack profile is calculated. The damage value of the arc contact image to be tested is calculated through the crack distribution density, which is used as the fatigue degree of the arc contact in the arc contact image to be tested.

[0017] Preferably, the tensile strength of the arc contact under test is obtained through the fatigue degree, and the specific operation steps are as follows:

[0018] A preset arc contact fatigue threshold e is set; it is then determined whether the fatigue degree of the arc contact in the image to be detected is greater than the arc contact fatigue threshold e.

[0019] If not, the tensile load of the arc contact in the image of the arc contact to be detected is deemed to be qualified, and the tensile strength of the arc contact is deemed to be qualified.

[0020] If so, the tensile load of the arc contact in the image of the arc contact to be tested is determined to be unqualified, and the tensile strength of the arc contact is determined to be unqualified.

[0021] Preferably, the strain field deviation value is calculated between the stable strain field and the speckle strain field, and the strain field deviation value is used to extract the strain field abrupt change region of the image of the arc contact to be detected. The specific operation steps are as follows:

[0022] Strain field distribution maps are plotted for the stable strain field and the speckle strain field respectively to obtain the stable strain field distribution map and the speckle strain field distribution map;

[0023] The maximum strain values ​​in the stable strain field distribution map and the speckle strain field distribution map are selected respectively; the strain field deviation value is obtained by calculating the maximum strain value in the stable strain field distribution map and the maximum strain value in the speckle strain field distribution map.

[0024] Establish a strain assignment threshold r, and determine whether the strain field deviation value is greater than the strain assignment threshold r by using the strain field deviation value;

[0025] If not, it is determined that the arc contact in the image of the arc contact to be detected does not show a sudden change in strain field;

[0026] If so, then the arc contact in the image of the arc contact to be detected shows a sudden change in the strain field;

[0027] Calculate the abrupt gradient for the image of the arc contact to be detected;

[0028] A Laplace operator is applied to the image of the arc contact to be detected to construct the stress and strain fields, thereby obtaining the local extrema of the stress and strain fields of the image of the arc contact to be detected, which can be used as potential abrupt change points of the stress and strain fields.

[0029] Construct the Hessian matrix of the stress and strain fields based on the second-order partial derivatives, and calculate the eigenvalues ​​of the Hessian matrix;

[0030] The gradient mutation points of the arc contact image to be detected are extracted by the mutation gradient, the potential mutation points, and the eigenvalues ​​of the Hessian matrix. All gradient mutation points are then set together to obtain the strain mutation region.

[0031] Preferably, the crack profile is extracted through the strain field abrupt change region; the crack distribution density of the crack profile is calculated, and the damage value of the arc contact image to be detected is calculated through the crack distribution density. The specific operation steps are as follows:

[0032] The strain abrupt change region is grown using a region growing algorithm to obtain the crack profile within the strain abrupt change region by performing gradient abrupt change point region growing.

[0033] The crack distribution density of the crack profile is calculated based on its extension direction; the surface morphology evolution of the image sequence is tracked using the crack distribution density to obtain a damage evolution image sequence; key damage feature parameters are extracted from the damage evolution image sequence to calculate the damage value of the arc contact in the image to be detected.

[0034] The key damage characteristic parameters include: crack distribution density, crack propagation rate, and crack morphology changes.

[0035] Preferably, the crack distribution density of the crack profile is calculated based on its extension direction; the surface morphology evolution is tracked using the crack distribution density to obtain a damage evolution image sequence. The specific operation steps are as follows:

[0036] The strain abrupt change region is uniformly divided into grid units, and the number of cracks in the internal crack profile of each grid unit is calculated.

[0037] The crack distribution density is calculated based on the number of cracks in each grid unit;

[0038] The crack distribution density of the arc contact image to be detected in the image sequence is tracked by the image registration algorithm, and the relationship between the crack position changes at different time points in the image sequence is established to obtain a damage evolution image sequence that tracks the surface morphology evolution.

[0039] Preferably, key damage feature parameters are extracted from the damage evolution image sequence, and the damage value of the arc contact in the image of the arc contact to be detected is calculated. The specific operation steps are as follows:

[0040] The total area of ​​the crack profile in the strain abrupt change region of each frame of the arc contact image to be detected is calculated in the damage evolution image sequence; the crack propagation rate is obtained by calculating the total area of ​​the crack profile between two consecutive frames.

[0041] The crack profile is analyzed using a skeletonization method to detect crack branch points, revealing that the crack profile shows branching and expansion, indicating a change in crack morphology.

[0042] The damage value of the arc contact in the image of the arc contact to be detected is calculated by the crack distribution density, crack propagation rate, and crack morphology changes.

[0043] Accordingly, the present invention also proposes a system for evaluating the tensile strength of arc contact images, characterized in that it includes: an acquisition module; a processing module; and an evaluation module;

[0044] The acquisition module is used to acquire tensile load image sequences of the arc contact, preprocess the tensile load image sequences of the arc contact to obtain the arc contact to be tested in the image sequence; the processing module is used to acquire the stress and strain field of the arc contact under tensile load in the image sequence of the arc contact to be tested; the fatigue degree of the arc contact to be tested is determined by the stress and strain field; the tensile strength of the arc contact to be tested is obtained by the fatigue degree; the evaluation module is used to generate an evaluation report of the arc contact based on the tensile strength of the arc contact to be tested.

[0045] Compared with the prior art, the embodiments of the present invention have at least the following technical advantages:

[0046] Analysis of the above-mentioned method and system for evaluating the tensile strength of an arc contact image provided by the present invention shows that, in specific applications, spraying uniform optical speckle onto the arc contact can analyze the deformation process of the arc contact under tensile load, and simultaneously record a high-definition image sequence under tensile load, and perform noise reduction preprocessing on the recorded high-definition image sequence to obtain the arc contact image to be tested.

[0047] Furthermore, for each image of the arc contact to be detected in the image sequence, pixels with the same gray value are clustered to help identify different regions in the image of the arc contact to be detected, especially those regions with consistent gray-level characteristics, such as surface speckle or uniform regions. The Euclidean distance between pixel clusters can be calculated to measure the distribution differences between different clusters. The dispersion of pixel clusters is obtained through the Euclidean distance between pixel clusters. If the dispersion between pixel clusters is large, it indicates that there are large differences in gray-level values ​​or textures within the region, which can reflect the tensile properties of the arc contact during tensile loading. The gray-level mean of each pixel cluster is calculated for speckle regions and stable regions in each image of the arc contact to be detected in the image sequence. The gray-level distribution in speckle regions is relatively chaotic, and selecting regions with higher gray-level mean pixel clusters may correspond to more prominent speckle features. The pixel cluster with the highest gray-level mean in the stable region is selected as the stable region. Feature points are identified; using optical flow, the stable feature points in the stable region of each arc contact image to be detected are sequenced for motion analysis, which yields accurate motion estimation and a stable displacement field of the stable feature points, reflecting the true deformation of the object. The stable displacement field directly reflects the motion state of the arc contact, while the strain field can further reveal the strain distribution of the material. Therefore, the stable displacement field is mapped to the strain field to obtain a stable strain field of stress and strain. The speckle region in the arc contact image to be detected is divided into multiple small blocks using a fast matching image registration method. The speckle displacement field is obtained through the image sequence of multiple small blocks. By dividing the speckle region into small blocks, the motion of the speckle region can be analyzed in more detail, thereby estimating the displacement of the speckle region more accurately. The speckle displacement field can also be mapped to obtain the strain field, revealing the stress and strain distribution on the surface of the arc contact, and helping to analyze the local overload or abnormal deformation that may occur in the arc contact during operation.

[0048] Furthermore, strain field distribution maps are plotted for both the stable strain field and the speckle strain field, resulting in stable strain field distribution maps and speckle strain field distribution maps. The strain field deviation value is calculated by selecting the maximum strain value in the stable strain field distribution map and the maximum strain value in the speckle strain field distribution map. If the deviation between the maximum strain value in the speckle strain field and the maximum strain value in the stable strain field is too large, it indicates abnormal strain change in that region. The abrupt change gradient is calculated on the image of the arc contact to be tested, and the abrupt change points of stress and strain fields are extracted using both the Laplacian operator and the Hessian matrix. The gradient abrupt change points are calculated using the abrupt change gradient and reflect the abrupt change region, thus obtaining the strain abrupt change region. A region growing algorithm is used to grow the gradient abrupt change region, obtaining the crack profile within the strain abrupt change region, which can identify the crack profile that may appear in the arc contact during tensile loading.

[0049] Furthermore, by querying the crack outline to determine the number of cracks, and calculating the crack distribution density based on the number of cracks, it is helpful to understand the specific distribution of cracks inside the material, thereby reflecting the tensile strength of the arc contact. The crack distribution density is used to track the evolution of all arc contact images to be tested, resulting in a damage evolution image sequence, which can show the distribution and morphology of cracks on the surface of the arc contact material in each frame of the arc contact image. The total area of ​​the crack outline in each frame of the arc contact image can be used to calculate the crack propagation rate, revealing the speed of crack propagation in each frame. Then, a skeletonization method is used to detect crack branch points on the crack outline, determining whether crack morphological changes have propagated, reflecting the tensile strength of the arc contact. Finally, the damage value of the arc contact in the image is calculated using the crack distribution density, crack propagation rate, and crack morphological changes. Attached Figure Description

[0050] Figure 1 This is a flowchart of the main process for evaluating the tensile strength of an arc contact image according to Embodiment 1 of the present invention.

[0051] Figure 2 This is a flowchart illustrating the process of identifying the tensile strength of an arc contact in an image of an arc contact to be detected, as part of an evaluation and processing method for the tensile strength of an arc contact image according to Embodiment 1 of the present invention.

[0052] Figure 3 This is a flowchart illustrating the calculation of damage values ​​of the arc contact in the image of the arc contact to be detected, according to a method for evaluating the tensile strength of an arc contact image in Embodiment 1 of the present invention.

[0053] Figure 4 This is a flowchart illustrating the damage value calculated based on key damage feature parameters in a method for evaluating the tensile strength of an arc contact image according to an embodiment of the present invention.

[0054] Figure 5 This is a flowchart of a system for evaluating the tensile strength of an arc contact image according to Embodiment 2 of the present invention;

[0055] Labels: Acquisition module 10; Processing module 20; Evaluation module 30. Detailed Implementation

[0056] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0057] Example 1

[0058] like Figure 1 As shown, the present invention provides a method for evaluating the tensile strength of an arc contact image, comprising the following steps:

[0059] S1: Acquire the tensile load image sequence of the arc contact, preprocess the tensile load image sequence of the arc contact to obtain the image sequence of the arc contact to be detected;

[0060] The surface of the arc contact is coated with a uniform optical speckle pattern;

[0061] It should be noted that, firstly, before acquiring the tensile load image of the arc contact, a layer of optical speckle is uniformly sprayed onto the surface of the arc contact for subsequent image analysis to obtain deformation information of the arc contact; the speckle pattern should be detailed and cover the key areas of the entire arc contact; a high-speed camera is used to capture the deformation process of the arc contact under tensile load in real time, and the tensile load image of the arc contact captured by the high-speed camera will be used for subsequent strain field analysis (that is, the high-speed camera is used to capture the deformation process of the arc contact in real time during tension or under stress, which can record high-frequency changes and help analyze the behavior characteristics of the arc contact under dynamic load).

[0062] Simultaneously, a noise reduction preprocessing operation is performed on the tensile load image of the arc contact to obtain the image of the arc contact to be detected;

[0063] S2: Obtain the stress and strain field of the arc contact under tensile load from the image sequence of the arc contact to be detected; determine the fatigue degree of the arc contact to be detected based on the stress and strain field; obtain the tensile strength of the arc contact to be detected based on the fatigue degree.

[0064] The stress and strain fields include speckle strain fields and steady strain fields;

[0065] It should be noted that the stress field describes the stress distribution at various points of an object under the action of external force; the stress and strain fields describe the deformation of an object due to stress. These fields can reveal fatigue cracks or other damaged areas that may exist on the surface of the arc contact. By analyzing the anomaly regions or local high strain regions in the stress and strain fields, the fatigue damage of the material can be inferred. These high strain regions are often associated with cracks and can be used as a basis for judging fatigue degree.

[0066] Meanwhile, stress and strain fields include speckle strain fields and steady-state strain fields. A steady-state strain field refers to the situation where, during the deformation of an object, the strain distribution inside or on the surface of the object tends to remain unchanged after the external force or strain reaches a certain stable state. In other words, a steady-state strain field describes the state where, after the deformation of an object reaches a certain stage, the object's strain field no longer changes with time or load, exhibiting a stable strain state. A speckle strain field refers to the method of measuring the strain distribution by analyzing the changes in the speckle pattern on the surface of an object. A speckle pattern is a small, randomly distributed pattern of light spots formed by the interference effect of light, usually formed by coating the surface of an object with tiny particles or using some other technology. By capturing the speckle pattern on the surface of an object before and after loading with a high-resolution camera and performing image processing (such as digital image correlation, DIC), the local strain distribution on the surface of the object during the deformation process can be obtained.

[0067] Under high fatigue conditions, the tensile strength of the arc contact may be weakened, meaning it may not be able to withstand greater loads and may even break.

[0068] The fatigue value obtained from stress and strain field analysis can be used to further estimate the tensile strength of the arc contact; this provides an important basis for the subsequent evaluation report.

[0069] S3: Generate an evaluation report for the arc contact based on its tensile strength;

[0070] It should be noted that the report usually includes the fatigue degree, tensile strength, and possible crack or damage areas of the arc contact. The remaining service life of the arc contact is assessed based on the fatigue degree, reflecting the health status assessment of the arc contact.

[0071] Specifically, such as Figure 2 As shown, in step S2, the stress and strain field of the tensile load of the arc contact to be detected are obtained from the image of the arc contact to be detected. The steps include:

[0072] For each image of the arc contact to be detected in the image sequence, pixels with the same gray value are clustered to obtain pixel clusters; the dispersion of the pixel clusters is analyzed, and speckle regions and stable regions in the image of the arc contact to be detected are obtained based on the dispersion of the pixel clusters; displacement field operations are performed on the speckle regions and stable regions to obtain speckle strain fields and stable strain fields. The specific operation steps are as follows:

[0073] S21: Cluster the pixels with the same gray value in each image of the arc contact to be detected in the image sequence to obtain pixel clusters;

[0074] Calculate the Euclidean distance between the pixel clusters, and obtain the dispersion of the pixel clusters through the Euclidean distance between the pixel clusters;

[0075] A preset distribution threshold t for pixel clusters is established; it is then determined whether the dispersion of the pixel clusters is greater than the distribution threshold t.

[0076] If not, the pixel clusters are determined to be unevenly dispersed, and the unevenly dispersed pixel clusters are taken as speckle regions.

[0077] If so, then the pixel clusters are determined to be uniformly dispersed, and the uniformly dispersed pixel clusters are considered as stable regions.

[0078] It should be noted that clustering pixels in an image according to their grayscale values ​​forms pixel clusters. The purpose of clustering is to group pixels in an image that have similar grayscale values ​​or similar features together. This can help identify different regions in an image, especially those regions with consistent grayscale features, such as surface speckle or uniform regions.

[0079] Calculate the Euclidean distance between pixel clusters. The Euclidean distance can measure the distribution difference between different clusters. The smaller the distance, the closer the clusters are, and the more uniform the gray values ​​or texture of the image area. The larger the distance, the greater the difference between clusters, and the more significant the texture change of the image area, thus reflecting the degree of dispersion of pixel clusters.

[0080] Based on the Euclidean distance between pixel clusters, the dispersion (i.e., the degree of dispersion of pixel clusters) is calculated. Dispersion reflects the distribution between pixel clusters. If the dispersion between pixel clusters is large, it indicates that there are significant differences in grayscale values ​​or textures within the region. It also reflects that during tensile loading of the arc contact, the stretched portion of the arc contact is affected by the dispersion of pixels, resulting in a larger dispersion between pixel clusters in the tensile-loaded portion of the arc contact. If the dispersion is small, it indicates that the grayscale values ​​or textures within the region are relatively uniform. This means that the pixel clusters in the unloaded portion or the tensile-loaded portion of the arc contact have not undergone significant deformation, reflecting good tensile properties of the arc contact during tensile loading. For example, in material testing, speckle areas may be related to surface defects or local deformation, while stable areas represent normal areas of the material.

[0081] By setting a preset pixel cluster distribution threshold t, the system filters out which pixel clusters in the image of the arc contact to be detected have a large dispersion, forming speckle areas (i.e., speckle areas represent areas where the arc contact may deform after being subjected to tensile load); and which pixel clusters have a small dispersion, forming stable areas (i.e., stable areas represent areas that have not undergone tensile load but have deformed, or areas that have undergone tensile load but have not shown any deformation after stretching).

[0082] S22: Calculate the average gray value of each pixel cluster in the speckle region and stable region of each arc contact image to be detected in the image sequence;

[0083] The pixel clusters with the highest and largest number of gray-scale values ​​in the speckle region are selected (e.g., if there are 10 pixel clusters in the speckle region, the highest gray-scale value is 200, and there are 2 such clusters; the second highest gray-scale value is 180, and there are 5 such clusters; in this case, the second highest gray-scale value of 180 is selected as the speckle feature point; first, the pixel clusters in the speckle region are sorted according to their gray-scale values, and the pixel clusters with the highest and largest number of gray-scale values ​​are selected as the speckle feature points. This can reflect the tensile deformation changes in different parts of the speckle region, thus better reflecting the tensile properties of the arc contact), as the speckle feature points.

[0084] Clusters of pixels with the highest gray-scale mean in the stable region are selected as stable feature points;

[0085] It should be noted that selecting the cluster with the highest gray-level mean in the speckle region means identifying the most significant gray-level part in that region. Typically, the gray-level distribution in speckle regions is quite chaotic; selecting areas with higher gray-level mean values ​​may correspond to more prominent speckle features, which helps extract key change information in subsequent image processing and analysis. Speckle feature points may represent noise, deformation, or local reflection points in the image, and are crucial for subsequent processing and change analysis.

[0086] In stable regions, selecting the cluster with the highest gray-scale mean means selecting the most representative and consistent part of that region. Stable regions typically reflect certain areas in an image that are unaffected by changes, and the gray-scale values ​​of these areas are usually relatively stable. By selecting the region with the highest gray-scale mean, we can effectively obtain the unchanging or most iconic parts of the image. Stable feature points are helpful for subsequent tracking and matching because these points can serve as reference benchmarks to help determine the deformation or positional changes of objects in the image sequence.

[0087] S23: Using optical flow, the stable feature points in the stable region of each arc contact image to be detected are subjected to image sequence motion to obtain the stable displacement field (i.e., motion vector) of the stable feature points. The stable displacement field refers to the situation where the displacement distribution of the surface or the interior of the material remains stable and uniform or does not change significantly in a certain direction during the deformation process of an object under force; in short, it means that the displacement distribution of the object tends to be stable over time and the displacement field is not affected by additional disturbances or changes.

[0088] The stable displacement field is mapped to the strain field to obtain the stable strain field of stress and strain fields;

[0089] It should be noted that optical flow is a method for estimating the displacement of an object or surface in an image by analyzing the motion of pixels in an image sequence. In this step, optical flow is used to track stable feature points in the arc contact image (these feature points typically remain stable throughout the image sequence) and calculate the displacement field (motion vector) of the stable region. These stable feature points appear as areas with small displacements or no disturbance in the image sequence, usually reflecting areas of the object surface that are relatively smooth or do not change significantly.

[0090] By performing displacement analysis on stable feature points in a stable region using the optical flow method, accurate motion estimation can be obtained, especially in areas on the arc contact surface where there is no significant deformation or noise interference. In this case, the motion of stable feature points reflects the true deformation of the object, and therefore can be used to infer the strain and stress changes on the arc contact surface.

[0091] The obtained stable displacement field is mapped to a strain field. The displacement field describes the movement of each stable feature point in the image of the arc contact to be detected, while the strain field describes the deformation of the object at different positions. It is usually calculated by the change of the displacement field. The deformation of each point on the surface of the arc contact is obtained by the stable displacement field (motion vector), which is the basis of stress-strain analysis.

[0092] The steady displacement field directly reflects the motion state of the arc contact, while the strain field can further reveal the strain distribution of the material. The purpose of this mapping is to obtain the physical deformation during the operation of the arc contact, which helps to understand and predict the performance of the arc contact under actual working conditions.

[0093] S24: The speckle region in the image of the arc contact to be detected is divided into multiple small blocks using a fast matching image registration method;

[0094] Using the speckle feature point as the center, a small patch around the speckle feature point is randomly selected as a location block; the location block is then filtered out by traversing the speckle region in all the arc contact images to be detected in the image sequence.

[0095] By using the mean squared difference of grayscale values ​​of all location blocks, the two location blocks with the highest similarity are selected.

[0096] The offset of the speckle feature points in the speckle region of the image of the arc contact to be detected in the two position blocks is calculated to obtain the displacement vector of the speckle feature points, which is used as the speckle displacement field of the speckle feature points (i.e., the speckle displacement field refers to the use of surface speckle patterns (patterns composed of tiny, randomly distributed points or spots) to track the displacement of the object surface during deformation. The speckle pattern is usually formed by spraying optical speckle (such as fine particles) onto the object surface. The speckle pattern of the object surface before and after loading is captured by a high-speed camera, and the changes in the pattern are compared by image processing technology to calculate the displacement of each point on the surface).

[0097] The speckle displacement field is mapped to the strain field to obtain the speckle strain field of stress and strain fields;

[0098] It should be noted that in the image of the arc contact to be detected, the speckle region is divided into multiple small blocks, and the small blocks are selected for matching based on the position of the speckle feature points. The block matching method is usually used for image registration, which finds matching points in the speckle region by comparing the similarity (such as the mean square difference of gray levels) of each block.

[0099] The speckle region contains local texture information of the image of the arc contact to be detected. These textures can provide information for accurate registration and displacement estimation. By dividing the speckle region into small blocks, the motion of the speckle region can be analyzed in more detail, thereby estimating the displacement of the speckle region more accurately. This method is particularly suitable for processing high-resolution image data.

[0100] By calculating the mean square difference of gray levels between different location blocks, the location block with the highest similarity is selected, and the displacement is estimated by calculating the offset between these two location blocks; this is a matching method based on similarity calculation, which can help accurately find the motion trajectory of speckle feature points.

[0101] Speckle feature points are usually represented by similar grayscale patterns in different images of the arc contact to be detected. Using the grayscale mean square error as the matching standard helps to reduce the interference of factors such as illumination changes and noise, and ensures the consistency and accuracy of the selected matching blocks in different images.

[0102] By calculating the offset of the two selected location blocks, the displacement vector of the speckle feature points is obtained; the speckle displacement field is the set of displacement vectors of these speckle feature points, representing the displacement of the speckle region in the image sequence.

[0103] Speckle feature points are important local feature points in the image of the arc contact to be detected. Their movement can reflect the displacement of a certain part of the arc contact image. By calculating the displacement vectors of these speckle feature points, more refined motion information can be obtained, and then the strain field of the object surface can be derived. This process is crucial for accurately describing the motion and deformation of the arc contact. Similar to the stable displacement field in step S23, the speckle displacement field can also be mapped to obtain the strain field, revealing the stress and strain distribution on the surface of the arc contact.

[0104] The speckle region is a local texture point in the image of the arc contact to be detected. Their displacement changes can provide more detailed information, especially in the local area of ​​the arc contact. Mapping the speckle displacement field to a strain field helps to reveal local deformation and helps to analyze the local overload or abnormal deformation that may occur in the arc contact during operation.

[0105] In step S2, the specific operational steps for the step "calculating the fatigue degree of the arc contact to be tested using the stress and strain field; obtaining the tensile strength of the arc contact to be tested using the fatigue degree" are as follows:

[0106] S25: Calculate the strain field deviation value between the stable strain field and the speckle strain field, extract the strain field abrupt change region of the arc contact image to be tested using the strain field deviation value, extract the crack profile through the strain field abrupt change region; calculate the crack distribution density of the crack profile, calculate the damage value of the arc contact image to be tested using the crack distribution density, and use it as the fatigue degree of the arc contact in the arc contact image to be tested;

[0107] It should be noted that the strain field describes the strain change of an object's surface under the action of external forces; the steady strain field usually refers to the relatively uniform strain distribution on the surface of an object under load; while the speckle strain field refers to the region of abrupt strain change caused by local damage, cracks or other defects.

[0108] By calculating the deviation between the stable strain field and the speckle strain field, local strain anomaly regions on the surface can be found, that is, strain field abrupt change regions, which are usually manifestations of cracks or fatigue damage.

[0109] Since cracks typically cause significant changes in the strain field, extracting the crack profile from abrupt regions can help identify the location and morphology of damage on the material surface.

[0110] Calculating crack distribution density helps to quantify the degree of damage to materials. Dense crack distribution usually means that the material has more severe fatigue damage, which may affect its performance and lifespan.

[0111] The damage value calculated based on the crack distribution density of the crack profile is used as the fatigue degree, which reflects the damage level of the arc contact. The higher the fatigue degree, the more severe the fatigue damage experienced by the material, which may lead to earlier failure.

[0112] S26: Preset arc contact fatigue threshold e; determine whether the fatigue of the arc contact in the image to be detected is greater than the arc contact fatigue threshold e;

[0113] If not, the tensile load of the arc contact in the image of the arc contact to be detected is deemed to be qualified, and the tensile strength of the arc contact is deemed to be qualified.

[0114] If so, the tensile load of the arc contact in the image of the arc contact to be detected is determined to be unqualified, and the tensile strength of the arc contact is determined to be unqualified.

[0115] It should be noted that the preset fatigue threshold ee, which is a critical fatigue value, is used to determine whether the arc contact meets the tensile strength requirements. This threshold represents the maximum fatigue damage that the material can withstand. When the fatigue exceeds this threshold, it indicates that the fatigue damage of the arc contact is too great.

[0116] If the fatigue degree of the arc contact under test is below the threshold, the tensile strength of the arc contact is deemed to be qualified, meaning that the arc contact can safely withstand tensile loads and will not fail; if the fatigue degree is above the threshold, it indicates that the fatigue damage of the arc contact is too great, the tensile strength is unqualified, and there is a high risk of failure.

[0117] Specifically, such as Figure 3 As shown, in step S25, the strain field deviation value is calculated between the stable strain field and the speckle strain field. The strain field deviation value is used to extract the strain field abrupt change region of the arc contact image to be detected, and the crack profile is extracted through the strain field abrupt change region. The crack distribution density of the crack profile is calculated, and the damage value of the arc contact image to be detected is calculated through the crack distribution density. The specific operation steps are as follows:

[0118] S251: Draw strain field distribution diagrams for the stable strain field and the speckle strain field respectively to obtain the stable strain field distribution diagram and the speckle strain field distribution diagram;

[0119] The maximum strain values ​​in the stable strain field distribution map and the speckle strain field distribution map are selected respectively; the strain field deviation value is obtained by calculating the maximum strain value in the stable strain field distribution map and the maximum strain value in the speckle strain field distribution map.

[0120] Establish a strain assignment threshold r, and determine whether the strain field deviation value is greater than the strain assignment threshold r by using the strain field deviation value;

[0121] If not, it is determined that the arc contact in the image of the arc contact to be detected does not show a sudden change in strain field;

[0122] If so, then the arc contact in the image of the arc contact to be detected shows a sudden change in the strain field;

[0123] It should be noted that a steady strain field refers to the strain distribution of an arc contact under normal operating conditions without external disturbances or material defects, representing the deformation of a material or structure under ideal conditions; a speckle strain field is usually a strain field obtained through optical or other sensor technologies, which may cause local fluctuations in strain due to experimental errors, noise, surface speckle, and other factors.

[0124] Comparing these two can reveal the difference between noise caused by speckle and normal strain field; in particular, if the speckle strain field in some areas is much larger than the strain value of the stable strain field, it may indicate that strain abrupt change has occurred in that area, that is, the material or structure may have abnormal deformation or damage in that area.

[0125] The deviation value reflects the difference between the speckle strain field and the steady-state strain field. If the deviation between the maximum strain value of the speckle strain field and the maximum strain value of the steady-state strain field is too large, it indicates that the strain change in this region is abnormal, which may be due to abrupt changes caused by external factors. A threshold r is set to determine whether the strain field deviation is greater than the threshold. The purpose of this is to avoid false abrupt results caused by small measurement errors or noise. If the deviation value exceeds the threshold r, it is considered that a strain field abrupt change has occurred in this region, indicating possible damage, stress concentration or structural problems.

[0126] S252: Calculate the abrupt gradient for the image of the arc contact to be detected;

[0127] A Laplace operator is applied to the image of the arc contact to be detected to construct the stress and strain fields, thereby obtaining the local extrema of the stress and strain fields of the image of the arc contact to be detected, which can be used as potential abrupt change points of the stress and strain fields.

[0128] Construct the Hessian matrix of stress and strain fields based on the second-order partial derivatives, and calculate the eigenvalues ​​of the Hessian matrix (i.e., the eigenvalues ​​of the Hessian matrix can reveal abrupt change points, which are usually located in the local extreme regions of the strain field and may indicate local instability or crack initiation points of the material).

[0129] The gradient mutation points of the arc contact image to be detected are extracted by the mutation gradient, the potential mutation points and the eigenvalues ​​of the Hessian matrix. All gradient mutation points are set together to obtain the strain mutation region (that is, the gradient mutation point set refers to those points with large gradient changes (i.e. mutation gradient), significant Laplacian operator values, or abnormal changes in the eigenvalues ​​of the Hessian matrix. These points are the potential strain mutation regions).

[0130] It should be noted that the abrupt change gradient represents the rate of local change in the strain field. Generally, a region with a large abrupt change gradient indicates a drastic change in strain or stress at that location, which may be a precursor to crack propagation or other anomalies in the material; therefore, calculating the abrupt change gradient can help identify these potential regions of abrupt strain change.

[0131] In a strain field, the Laplace operator can highlight local extrema, i.e., the points where strain changes most drastically; abrupt change regions are usually located at local extrema of strain, and these points may correspond to local instability of the material, crack initiation, or other structural problems; local extrema obtained by the Laplace operator can serve as potential abrupt change points in stress and strain fields, further improving the accuracy of identifying abrupt change regions.

[0132] The Hessian matrix is ​​a matrix composed of second-order partial derivatives, used to describe the local curvature of a function (such as a strain field) at a certain point. The eigenvalues ​​of the Hessian matrix can reveal the direction and intensity of the changes in the strain field. If there are outliers in the eigenvalues ​​of the Hessian matrix, it indicates that there are large changes in the strain or stress field in that region. These regions usually correspond to local extreme points or stress concentration points, which may lead to crack propagation or other instability phenomena.

[0133] By combining abrupt gradients, local extrema, and eigenvalues ​​of the Hessian matrix, gradient abrupt points in the strain field can be accurately identified. These gradient abrupt points are markers of abrupt change regions, and the drastic strain changes that usually occur at these locations are caused by local stress concentrations, material defects, or other structural anomalies.

[0134] The strain abrupt change region is composed of these gradient abrupt change points and is a potential risk area; by extracting these regions, abnormal areas (i.e. crack profiles) that may exist in the arc contact can be identified in advance.

[0135] S253: Use a region growing algorithm to grow the strain abrupt change region at gradient abrupt change points to obtain the crack profile within the strain abrupt change region.

[0136] It should be noted that the region growing algorithm is common knowledge. One gradient mutation point is selected as a seed point for neighborhood growing, and finally the crack profile is obtained. This will not be elaborated here. During the tensile loading process, cracks may appear in the arc contact. Therefore, the crack profile is obtained by performing region growing through the abnormal gradient protrusion point, and then the crack distribution density is calculated through the crack profile.

[0137] S254: Calculate the crack distribution density of the crack profile through the extension direction of the crack profile; track the surface morphology evolution of the image sequence through the crack distribution density to obtain a damage evolution image sequence; extract key damage feature parameters through the damage evolution image sequence to calculate the damage value of the arc contact image to be detected.

[0138] The key damage characteristic parameters include: crack distribution density, crack propagation rate, and crack morphology changes;

[0139] It should be noted that the crack profile is obtained through the region growing algorithm, and the crack distribution density is calculated based on the extension direction, distribution and number of the crack profile. The crack density distribution can reflect the severity of fatigue damage on the material surface; the denser the crack distribution, the more severe the fatigue damage may be.

[0140] As the time-series images (i.e., image sequences) evolve, the crack distribution becomes increasingly dense, and the morphology of the arc contact surface changes. Larger cracks, fatigue pits, or other damage morphologies may gradually appear on the surface. By comparing and analyzing a series of time-series images, the evolution of the surface morphology can be tracked, allowing us to understand the damage evolution of the material under different fatigue cycles. Through time-series image data, a series of images depicting the evolution of damage on the arc contact surface can be established. Each image represents the damage state of the arc contact surface at a certain time point or under a fatigue cycle, thereby calculating the damage value.

[0141] Specifically, such as Figure 4 As shown, in step S254, the crack distribution density of the crack profile is calculated based on its extension direction; the surface morphology evolution is tracked using the crack distribution density to obtain a damage evolution image sequence. The specific operation steps are as follows:

[0142] S2541: The strain abrupt change region is uniformly divided into grid units, and the number of cracks in the internal crack profile of each grid unit is calculated.

[0143] The crack distribution density is calculated based on the number of cracks in each grid unit;

[0144] It should be noted that the strain abrupt change region is divided into multiple small grid units, each of which is a computational region used to analyze the crack characteristics within that region. The number of cracks in the crack profile within each grid unit is calculated. By calculating the number of cracks in each grid unit, the crack density of the region can be quantified. This helps to understand the specific distribution of cracks within the material. In the strain abrupt change region, cracks may have a high density, affecting the mechanical properties of the material. Therefore, accurately calculating the number of cracks helps to assess the degree of damage in this region.

[0145] S2542: Using an image registration algorithm, the crack distribution density of the arc contact image to be detected in the image sequence is tracked, the crack position change relationship at different time points in the image sequence is established, and a damage evolution image sequence tracking the surface morphology evolution is obtained.

[0146] It should be noted that image registration algorithms (such as feature-based registration, phase correlation, etc.) are used to align the images of the arc contact to be detected at different time points (i.e., the image sequence is a time series image, so it has images of the arc contact to be detected at different time points), thereby realizing crack tracking. This allows the relationship between crack position changes at different time points to be established, and then the dynamic process of crack propagation can be analyzed. By analyzing the crack distribution density and tracking the crack propagation trajectory, a damage evolution image sequence can be obtained. Each frame of the arc contact to be detected shows the distribution and morphology of cracks on the surface of the arc contact material at a specific time or under a specific loading condition.

[0147] In step S254, the following steps were performed on the step "extracting key damage feature parameters from the damage evolution image sequence and calculating the damage value of the arc contact in the image of the arc contact to be detected":

[0148] S2543: Calculate the total area of ​​the crack profile in the strain abrupt change region of each frame of the arc contact image to be detected for the damage evolution image sequence (i.e., the calculation of the total area of ​​the crack profile is common knowledge and will not be elaborated further).

[0149] The crack propagation rate is calculated by setting the time interval Δt and using the total area of ​​the crack profile between two consecutive frames. The calculation formula is as follows:

[0150]

[0151] In the formula, v crack Crack propagation rate is represented by the crack distribution density in the damage evolution image sequence.

[0152] A1 represents the total area of ​​the crack profile in the previous frame of the damage evolution image sequence (i.e., the total area of ​​the crack profile at time t).

[0153] A2 represents the total area of ​​the crack profile in the next frame of the damage evolution image sequence (i.e., the total area of ​​the crack profile at time t-1).

[0154] It should be noted that the crack propagation rate is obtained through this formula, which describes the speed at which the crack propagates over time. A higher propagation rate indicates that the crack propagates faster within that time period, while a lower rate indicates that the crack propagates slower. At the same time, it can also reflect that the damage to the arc contact under tensile load is gradually increasing. Therefore, the crack propagation rate can also briefly indicate the duration of the arc contact's tensile resistance, thus reflecting the tensile resistance of the arc contact.

[0155] S2544: Use the skeletonization method to detect crack branch points in the crack profile, and determine whether the crack profile branches in different time frames.

[0156] If not, it is determined that the crack profile has not branched (that is, by judging the crack profile at different time frames, if the crack profile has not branched, it means that the crack profile may not have changed significantly over time, and the arc contact has good tensile strength).

[0157] If so, it is determined that the crack profile has branched and expanded, and the crack profile has changed in crack morphology;

[0158] It should be noted that applying skeletonization to the crack profile makes it easier to detect the crack's topological structure, such as whether there are complex changes like branching or merging. Skeletonization is an image processing technique that transforms the crack profile in an image into a slender "skeleton" shape, thus simplifying the crack's structure. In the crack skeleton image, the branch points are where the crack splits from a main branch into multiple sub-branches. By analyzing the topological structure of the crack skeleton, these branch points can be identified.

[0159] By comparing the crack skeleton at different time frames, it can be determined whether the crack has branched and expanded. If the crack skeleton shows a splitting from a main crack into multiple branches in two consecutive frames, it can be determined that the crack has branched and expanded.

[0160] If crack branches appear, it indicates that the crack morphology has changed and the propagation is more complex. This usually means that the crack development has entered a new stage and may lead to more serious damage. In this case, the crack morphology has undergone a "morphological change".

[0161] S2545: The damage value of the arc contact in the image of the arc contact to be detected is calculated based on the crack distribution density, crack propagation rate, and crack morphology changes. The calculation formula is as follows:

[0162]

[0163] In the formula, ρ(xy) represents the crack distribution density within each grid unit.

[0164] exp(-μt) represents the time diffusion factor (i.e., the time dependence of crack propagation rate, which means that as time goes by, the crack propagation rate in the damage evolution image sequence will become faster and faster, and the crack speed will also become faster and faster).

[0165] α represents the weighting coefficient of crack distribution density (i.e., the contribution used to adjust the density term);

[0166] dA is the area of ​​a micro-element (which represents the integral over the entire strain abrupt change region. The micro-element area is the basic unit for integral calculation and is used to refine the local manifestation of crack propagation. By summing at the micro-element scale, the crack propagation situation of the entire region can be obtained).

[0167] v crack Expressed as crack propagation rate;

[0168] cos 2 θ represents the crack directionality factor (i.e., θ is the angle between the crack propagation direction and the principal stress direction);

[0169] k(s) represents the curvature function (which describes the degree of curvature of the crack path, indicating the complexity of the crack branching direction);

[0170] β represents the weighting coefficient for crack propagation rate;

[0171] i represents one of the branch directions of the crack;

[0172] M(s) represents the crack morphology factor (i.e., the crack shape characteristics of crack morphology changes);

[0173] ψ(s) represents the crack morphology function (i.e., the characteristic of the crack profile);

[0174] γ represents the weighting coefficient of the crack morphology factor;

[0175] ds is a line integral infinitesimal element along the crack boundary (i.e., the crack is usually a continuous boundary, and integration along the boundary can obtain various parameters on the crack boundary (such as crack propagation rate, curvature, etc.); by performing line integration along the crack boundary, the crack propagation and its impact on the material can be accurately calculated).

[0176] It should be noted that the above formula actually transforms the incremental accumulation process into a "state description"; the influence of time history is implicitly included through the exp(-μt) term; at the same time, the integrals of each term essentially contain the cumulative effect of damage evolution; and the current damage level (i.e., ρ(xy) represents the crack distribution density in each grid unit) is compared with v. crackThe damage development trend and the complexity of the damage represented by κ(s) and ψ(s) reflect the total damage value of the arc contact. This avoids the complexity of stepwise accumulation calculation (i.e., calculating the incremental damage value of the arc contact, which is represented by first calculating the damage value of two adjacent frames of the arc contact image to be detected, and then obtaining the total damage value of the arc contact through the damage value of each frame. However, this calculation method avoids accumulation calculation through the factors and other parameters in the above formula), reduces the accumulation of calculation error, and improves calculation efficiency.

[0177] The crack distribution density, crack propagation rate, crack morphology, and various other factors (such as directionality, curvature, and shape) all have different effects on crack propagation. By weighting and adjusting the contributions of these factors, the damage value of the arc contact was calculated.

[0178] Example 2

[0179] like Figure 5 As shown, the present invention also provides a system for evaluating the tensile strength of an arc contact image, comprising: an acquisition module 10; a processing module 20; and an evaluation module 30.

[0180] The acquisition module 10 is used to acquire tensile load image sequences of the arc contact, and preprocess the tensile load image sequences of the arc contact to obtain the arc contact to be detected in the image sequence.

[0181] The processing module 20 is used to acquire the stress and strain field of the tensile load of the arc contact in the image sequence; to determine the fatigue degree of the arc contact under test through the stress and strain field; and to obtain the tensile strength of the arc contact under test through the fatigue degree.

[0182] The evaluation module 30 is used to generate an evaluation report for the arc contact based on the tensile strength of the arc contact to be tested.

[0183] In summary, the present invention provides a method and system for evaluating the tensile strength of an arc contact image. This method analyzes the deformation process of the arc contact under tensile load by spraying a uniform optical speckle pattern onto the arc contact and simultaneously recording a high-definition image sequence under tensile load. The recorded high-definition image sequence is then pre-processed to remove noise, resulting in the arc contact image to be tested.

[0184] Furthermore, for each image of the arc contact to be detected in the image sequence, pixels with the same gray value are clustered to help identify different regions in the image of the arc contact to be detected, especially those regions with consistent gray-level characteristics, such as surface speckle or uniform regions. The Euclidean distance between pixel clusters can be calculated to measure the distribution differences between different clusters. The dispersion of pixel clusters is obtained through the Euclidean distance between pixel clusters. If the dispersion between pixel clusters is large, it indicates that there are large differences in gray-level values ​​or textures within the region, which can reflect the tensile properties of the arc contact during tensile loading. The gray-level mean of each pixel cluster is calculated for speckle regions and stable regions in each image of the arc contact to be detected in the image sequence. The gray-level distribution in speckle regions is relatively chaotic, and selecting regions with higher gray-level mean pixel clusters may correspond to more prominent speckle features. The pixel cluster with the highest gray-level mean in the stable region is selected as the stable region. Feature points are identified; using optical flow, the stable feature points in the stable region of each arc contact image to be detected are sequenced for motion analysis, which yields accurate motion estimation and a stable displacement field of the stable feature points, reflecting the true deformation of the object. The stable displacement field directly reflects the motion state of the arc contact, while the strain field can further reveal the strain distribution of the material. Therefore, the stable displacement field is mapped to the strain field to obtain a stable strain field of stress and strain. The speckle region in the arc contact image to be detected is divided into multiple small blocks using a fast matching image registration method. The speckle displacement field is obtained through the image sequence of multiple small blocks. By dividing the speckle region into small blocks, the motion of the speckle region can be analyzed in more detail, thereby estimating the displacement of the speckle region more accurately. The speckle displacement field can also be mapped to obtain the strain field, revealing the stress and strain distribution on the surface of the arc contact, and helping to analyze the local overload or abnormal deformation that may occur in the arc contact during operation.

[0185] Furthermore, strain field distribution maps are plotted for both the stable strain field and the speckle strain field, resulting in stable strain field distribution maps and speckle strain field distribution maps. The abrupt change gradient is calculated on the image of the arc contact to be tested, and the abrupt change points of stress and strain fields are extracted using both the Laplacian operator and the Hessian matrix. The gradient abrupt change points are calculated using the abrupt change gradient and the abrupt change points, reflecting the abrupt change region, thus obtaining the strain abrupt change region. A region growing algorithm is then used to grow the gradient abrupt change region, obtaining the crack profile within the strain abrupt change region, which can identify the crack profile that may appear in the arc contact during tensile loading.

[0186] Furthermore, by querying the crack profile to find the number of cracks, and calculating the crack distribution density based on the number of cracks, it is helpful to understand the specific distribution of cracks inside the material, thereby reflecting the tensile strength of the arc contact. Finally, an evaluation report of the arc contact is generated based on the tensile strength of the arc contact to be tested.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; those skilled in the art can modify the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the tensile strength of an arc contact image, characterized in that: The following steps are included: Acquire a tensile load image sequence of the arc contact, preprocess the tensile load image sequence of the arc contact, and obtain the image sequence of the arc contact to be detected; The surface of the arc contact is coated with a uniform optical speckle pattern; The stress and strain fields of the arc contact under tensile load are obtained from the images of the arc contact to be detected in the image sequence; the fatigue degree of the arc contact to be detected is determined by the stress and strain fields; and the tensile strength of the arc contact to be detected is obtained by the fatigue degree. The stress and strain fields include speckle strain fields and steady strain fields; An evaluation report for the arc contact is generated based on the tensile strength of the arc contact to be tested; To obtain the stress and strain field of the arc contact under tensile load from the image of the arc contact to be detected, the following steps are included: For each image of the arc contact to be detected in the image sequence, the pixels with the same gray value are clustered to obtain pixel clusters; The dispersion of the pixel clusters is analyzed, and the speckle region and stable region in the image of the arc contact to be detected are obtained based on the dispersion of the pixel clusters. Displacement field operations are performed on the speckle region and the stable region to obtain the speckle strain field and the stable strain field; The fatigue degree of the arc contact under test is calculated using the stress and strain fields. The specific operation steps are as follows: The strain field deviation value is calculated between the stable strain field and the speckle strain field. The strain field deviation value is used to extract the strain field abrupt change region of the arc contact image to be tested. The crack profile is extracted through the strain field abrupt change region. The crack distribution density of the crack profile is calculated. The damage value of the arc contact image to be tested is calculated through the crack distribution density, which is used as the fatigue degree of the arc contact in the arc contact image to be tested. The tensile strength of the arc contact under test is obtained by measuring the fatigue level. The specific operating steps are as follows: A preset arc contact fatigue threshold e is set; it is then determined whether the fatigue degree of the arc contact in the image to be detected is greater than the arc contact fatigue threshold e. If not, the tensile load of the arc contact in the image of the arc contact to be detected is deemed to be qualified, and the tensile strength of the arc contact is deemed to be qualified. If so, the tensile load of the arc contact in the image of the arc contact to be detected is determined to be unqualified, and the tensile strength of the arc contact is determined to be unqualified. The strain field deviation value is calculated between the stable strain field and the speckle strain field. The strain field abrupt change region of the image of the arc contact to be detected is extracted using the strain field deviation value. The specific operation steps are as follows: Strain field distribution maps are plotted for the stable strain field and the speckle strain field respectively to obtain the stable strain field distribution map and the speckle strain field distribution map; The maximum strain values ​​in the stable strain field distribution map and the speckle strain field distribution map are selected respectively; the strain field deviation value is obtained by calculating the maximum strain value in the stable strain field distribution map and the maximum strain value in the speckle strain field distribution map. Establish a strain assignment threshold r, and determine whether the strain field deviation value is greater than the strain assignment threshold r by using the strain field deviation value; If not, it is determined that the arc contact in the image of the arc contact to be detected does not show a sudden change in strain field; If so, then the arc contact in the image of the arc contact to be detected shows a sudden change in the strain field; Calculate the abrupt gradient for the image of the arc contact to be detected; A Laplace operator is applied to the image of the arc contact to be detected to construct the stress and strain fields, thereby obtaining the local extrema of the stress and strain fields of the image of the arc contact to be detected, which can be used as potential abrupt change points of the stress and strain fields. Construct the Hessian matrix of the stress and strain fields based on the second-order partial derivatives, and calculate the eigenvalues ​​of the Hessian matrix; The gradient mutation points of the arc contact image to be detected are extracted by the mutation gradient, the potential mutation points, and the eigenvalues ​​of the Hessian matrix. All gradient mutation points are then set together to obtain the strain mutation region.

2. The method for evaluating the tensile strength of an arc contact image according to claim 1, characterized in that: Crack profiles are extracted from the strain field abrupt change regions; the crack distribution density of the crack profiles is calculated, and the damage value of the arc contact image to be detected is calculated using the crack distribution density. The specific operation steps are as follows: The strain abrupt change region is grown using a region growing algorithm to obtain the crack profile within the strain abrupt change region by performing gradient abrupt change point region growing. The crack distribution density of the crack profile is calculated based on its extension direction; the surface morphology evolution of the image sequence is tracked using the crack distribution density to obtain a damage evolution image sequence; key damage feature parameters are extracted from the damage evolution image sequence to calculate the damage value of the arc contact in the image to be detected. The key damage characteristic parameters include: crack distribution density, crack propagation rate, and crack morphology changes.

3. The method for evaluating the tensile strength of an arc contact image according to claim 2, characterized in that: The crack distribution density of the crack profile is calculated based on its extension direction; the surface morphology evolution is tracked using the crack distribution density to obtain a damage evolution image sequence. The specific operation steps are as follows: The strain abrupt change region is uniformly divided into grid units, and the number of cracks in the internal crack profile of each grid unit is calculated. The crack distribution density is calculated based on the number of cracks in each grid unit; The crack distribution density of the arc contact image to be detected in the image sequence is tracked by the image registration algorithm, and the relationship between the crack position changes at different time points in the image sequence is established to obtain a damage evolution image sequence that tracks the surface morphology evolution.

4. The method for evaluating the tensile strength of an arc contact image according to claim 3, characterized in that: Key damage feature parameters are extracted from the damage evolution image sequence, and the damage value of the arc contact in the image of the arc contact to be detected is calculated. The specific operation steps are as follows: The total area of ​​the crack profile in the strain abrupt change region of each frame of the arc contact image to be detected is calculated in the damage evolution image sequence; the crack propagation rate is obtained by calculating the total area of ​​the crack profile between two consecutive frames. The crack profile was analyzed using a skeletonization method to detect crack branch points, resulting in branching and expansion of the crack profile, and a change in crack morphology. The damage value of the arc contact in the image of the arc contact to be detected is calculated by the crack distribution density, crack propagation rate, and crack morphology changes.

5. A system for evaluating the tensile strength of an arc contact image, used to implement the method for evaluating the tensile strength of an arc contact image according to any one of claims 1-4, characterized in that: include: Data acquisition module; Processing module; Evaluation module; The acquisition module is used to acquire tensile load image sequences of the arc contact, and preprocess the tensile load image sequences of the arc contact to obtain the arc contact to be detected from the image sequence. The processing module is used to acquire the stress and strain field of the tensile load of the arc contact in the image sequence; to determine the fatigue degree of the arc contact under test through the stress and strain field; and to obtain the tensile strength of the arc contact under test through the fatigue degree. The evaluation module is used to generate an evaluation report for the arc contact based on its tensile strength.

Images