Optical cable defect detection method and system based on image recognition model

Abstract

The invention relates to the technical field of computer vision, in particular to an optical cable defect detection method based on an image recognition model. The method comprises the following steps: acquiring an optical cable laying image; identifying an optical cable laying form according to the optical cable laying image to obtain exposed section optical cable data and folded section optical cable data; evaluating corrosion resistance of the optical cable based on exposed section optical cable data; determining the rusting degree of the armor layer according to the corrosion resistance of the optical cable; detecting optical fiber structure embrittlement according to the rusting degree of the armor layer to obtain optical fiber structure embrittlement data; enhancing the structural stability of the exposed section optical cable data according to the optical fiber structure embrittlement data to obtain structural stability data; constructing an optical cable defect image recognition model based on the optical fiber structure embrittlement data; and according to the optical cable defect image recognition model, detecting the microbend loss of the folding section optical cable data to obtain microbend loss data. According to the invention, the accuracy of optical cable defect identification is improved based on the computer vision technology.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to an optical cable defect detection method and system based on an image recognition model. Background Art

[0002] Optical cable armor protects optical fibers from mechanical damage and is typically made of metal, such as steel wire, aluminum, or aluminized steel. Optical cable defect detection uses automated analysis of cable images to detect defects such as cracks, scratches, and corrosion. However, existing technologies also have significant drawbacks. They rely heavily on high-quality, large amounts of annotated data. Insufficient or poor-quality training data significantly reduces the accuracy and robustness of the model. Traditional methods are unable to effectively identify minor defects or damaged areas in complex cable structures or environmental variations, leading to missed or false detections. In particular, during image processing, environmental factors such as lighting variations, occlusion, and noise can affect image quality, reducing detection effectiveness. Furthermore, most existing models are unable to adapt to different types of optical cables or installation environments. Encountering new cable types or non-standard installation methods requires retraining, increasing system maintenance costs. Traditional image recognition methods ignore the physical and chemical properties of optical cables (such as corrosion resistance and fatigue damage), which are crucial to their long-term stability and lifespan. Relying solely on image data cannot comprehensively assess the health of optical cables. In large-scale optical cable detection, the real-time performance and computational efficiency of existing methods often become bottlenecks, especially when large amounts of data need to be processed, resulting in slow processing speed and high resource consumption, which limits their promotion and popularization in practical applications. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide an optical cable defect detection method and system based on an image recognition model to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for detecting optical cable defects based on an image recognition model comprises the following steps:

[0005] Step S1: Acquire an optical cable laying image; identify the optical cable laying form according to the optical cable laying image, and obtain exposed section optical cable data and folded section optical cable data;

[0006] Step S2: Evaluating the corrosion resistance of the optical cable based on the exposed section optical cable data; determining the degree of armor rust based on the optical cable corrosion resistance; detecting the embrittlement of the optical fiber structure based on the degree of armor rust to obtain optical fiber structure embrittlement data; enhancing the structural stability of the exposed section optical cable data based on the optical fiber structure embrittlement data to obtain structural stability data;

[0007] Step S3: constructing an optical cable defect image recognition model based on the optical fiber structure embrittlement data; detecting microbending loss of the folded section optical cable data according to the optical cable defect image recognition model to obtain microbending loss data; locating the breakpoint of the folded section optical cable data based on the microbending loss data; and identifying optical fiber cross-sectional structural features based on the breakpoint.

[0008] Step S4: Transmitting the optical fiber cross-sectional structural features and structural stability data to the optical cable defect image recognition model, detecting the optical cable structural defects in the optical cable laying image, and generating optical cable structural defect data.

[0009] By identifying the cable layout and extracting data from exposed and folded sections, this method can accurately distinguish between cables laid in different configurations, facilitating more sophisticated subsequent defect detection tasks. Secondly, the corrosion resistance of the cable is assessed based on the exposed cable data, thereby better predicting the degree of armor rust. Furthermore, the structural stability of the exposed cable data is enhanced by fiber structural embrittlement data, effectively improving the ability to identify long-term damage such as cable aging and corrosion during detection. Furthermore, the fiber structural embrittlement data is used to construct an optical cable defect image recognition model. This model can detect microbend losses in folded cable sections and accurately locate breakpoints, providing a reliable basis for early warning of cable failures. By combining the structural characteristics of the fiber cross-section at the breakpoint, internal defects in the cable can be comprehensively analyzed and identified. This data is then transferred to the image recognition model, enabling more efficient and accurate detection of cable structural defects. This process significantly reduces the problems of missed and false detections common in traditional detection methods. By considering the physical and chemical properties of the cable, it enhances the comprehensive assessment of the cable's health, thereby improving the accuracy and robustness of detection. In addition, the efficient performance of this method in large-scale optical cable detection can effectively avoid the waste of computing resources and improve processing speed, which will help to better promote and popularize it in practical applications.

[0010] Preferably, step S1 is specifically as follows:

[0011] Step S11: Acquire an optical cable laying image;

[0012] Step S12: removing low-frequency noise from the optical cable laying image to obtain a noise-reduced optical cable laying image;

[0013] Step S13: Detecting the edge of the optical cable based on the denoised image of the optical cable laying, wherein the upper and lower detection thresholds are set to 50 and 150, respectively, to obtain the edge of the optical cable, and extracting the candidate optical cable area of the denoised image of the optical cable laying according to the edge of the optical cable;

[0014] Step S14: performing an expansion operation on the optical cable candidate region using a preset 3x3 structure element to obtain an optical cable expansion region;

[0015] Step S15: performing an erosion operation on the optical cable candidate area using a preset 5x5 structure element to obtain an optical cable erosion area;

[0016] Step S16: Calculate the aspect ratio of the cable corrosion area. If the aspect ratio is greater than 2:1, obtain the exposed cable segment data. Calculate the curvature of the cable corrosion area. If the curvature radius is less than 10 cm, obtain the folded cable segment data.

[0017] By identifying the cable layout and extracting data from exposed and folded sections, this method can accurately distinguish between cables laid in different configurations, facilitating more sophisticated subsequent defect detection tasks. Secondly, the corrosion resistance of the cable is assessed based on the exposed cable data, thereby better predicting the degree of armor rust. Furthermore, the structural stability of the exposed cable data is enhanced by fiber structural embrittlement data, effectively improving the ability to identify long-term damage such as cable aging and corrosion during detection. Furthermore, the fiber structural embrittlement data is used to construct an optical cable defect image recognition model. This model can detect microbend losses in folded cable sections and accurately locate breakpoints, providing a reliable basis for early warning of cable failures. By combining the structural characteristics of the fiber cross-section at the breakpoint, internal defects in the cable can be comprehensively analyzed and identified. This data is then transferred to the image recognition model, enabling more efficient and accurate detection of cable structural defects. This process significantly reduces the problems of missed and false detections common in traditional detection methods. By considering the physical and chemical properties of the cable, it enhances the comprehensive assessment of the cable's health, thereby improving the accuracy and robustness of detection. In addition, the efficient performance of this method in large-scale optical cable detection can effectively avoid the waste of computing resources and improve processing speed, which will help to better promote and popularize it in practical applications.

[0018] Preferably, the evaluating the corrosion resistance of the optical cable in step S2 includes:

[0019] Extracting surface texture data of exposed optical cable segment data;

[0020] Calculating gradients based on surface texture data;

[0021] Calculate the local gradient change rate based on the gradient;

[0022] Determine the optical cable corrosion area of the optical cable laying image based on the local gradient change rate;

[0023] Identify the corrosion depth based on the corroded area of the optical cable;

[0024] Calculate the corrosion rate based on the corrosion area of the optical cable;

[0025] The corrosion resistance of the exposed section of optical cable is evaluated based on the corrosion depth and corrosion rate.

[0026] The present invention helps to capture subtle changes in the surface of the optical cable by extracting surface texture data of the exposed section of the optical cable data, providing basic information for the subsequent identification of corrosion areas. By calculating the gradient based on these surface texture data, it is possible to accurately capture the subtle structural changes on the surface of the optical cable, and further by calculating the local gradient change rate, the corrosion area on the surface of the optical cable can be finely analyzed. By determining the corrosion area of the optical cable in the optical cable laying image, the damaged part of the optical cable can be accurately located, providing a reliable basis for the subsequent corrosion depth identification and rate calculation. According to the corrosion depth identified according to the corrosion area, the degree of damage to the optical cable structure caused by corrosion can be effectively evaluated, and by calculating the corrosion rate, the long-term loss process of the optical cable can be better monitored, providing a scientific basis for the corrosion resistance evaluation of the optical cable. These steps can comprehensively consider the corrosion status of the surface and interior of the optical cable, ensure a more comprehensive and accurate assessment of the health status of the optical cable, and provide data support for fault prediction and maintenance decisions in practical applications, thereby improving the accuracy and efficiency of detection.

[0027] Preferably, determining the degree of rust of the armor layer in step S2 includes:

[0028] Corrosion simulation was performed based on the corrosion resistance of the optical cable, wherein the temperature range was set to 25°C-40°C and the humidity range was set to 30%RH-98%RH to obtain corrosion simulation data;

[0029] Calculate the thickness of the optical cable armor layer based on corrosion simulation data;

[0030] Calculate and obtain thickness reduction data using the preset optical cable armor layer benchmark thickness and the optical cable armor layer thickness;

[0031] The electrical conductivity of the corrosion simulation data was recorded using an eddy current detector;

[0032] Determine corrosion depth based on electrical conductivity;

[0033] The degree of rust of the armor layer is determined based on the corrosion depth and thickness reduction data.

[0034] The present invention performs corrosion simulation by setting the range of temperature and humidity, and can generate diversified corrosion data under different environmental conditions, thereby simulating the corrosion process of optical cables under different conditions of use. Next, the thickness of the optical cable armor layer is calculated and compared with the preset reference thickness to obtain the thinning data of the armor layer, reflecting the actual degree of corrosion. The conductivity data recorded by the eddy current detector can be used to further evaluate the corrosion depth of the armor layer. The conductivity is closely related to the corrosion depth. This data can be used to more accurately understand the actual corrosion situation of the optical cable during use. Finally, combining the corrosion depth and thickness thinning data, the degree of rust of the armor layer can be accurately determined, providing a comprehensive assessment of the corrosion resistance of the optical cable. These steps can not only accurately simulate the corrosion situation of the optical cable in a real environment, but also effectively combine physical and chemical properties to ensure the long-term stability evaluation of the optical cable, overcome the defects of traditional image recognition methods that rely on environmental changes and data quality, and at the same time improve detection accuracy and efficiency, which helps to achieve more effective optical cable maintenance and management in practical applications.

[0035] Preferably, the detecting of optical fiber structure embrittlement in step S2 includes:

[0036] Determine the rust area according to the rust degree of the armor layer;

[0037] Determine the degree of armor failure based on the rusted area;

[0038] Divide the exposed cable data into armored rusted cable segments based on the degree of armor failure;

[0039] The fiber bending load is simulated for the rusted armored cable section, and the fiber fatigue life is predicted to obtain the fiber fatigue life;

[0040] The optical fiber structure embrittlement is detected based on the optical fiber fatigue life to obtain the optical fiber structure embrittlement data.

[0041] By demarcating rusted areas on the armor layer, this method can accurately identify key sections of optical cables susceptible to corrosion in real-world environments, which is crucial for subsequent failure analysis. Based on the delineation of rusted areas, the degree of armor failure can be determined, thereby assessing the risk of damage to the cable during long-term use. Further segmenting the exposed cable data into sections with rusted armor provides an accurate basis for subsequent damage prediction and analysis. By subjecting these rusted sections to fiber bending load simulation, the fatigue process of the optical fiber can be simulated under actual application conditions, thereby predicting its fatigue life. This prediction provides a scientific basis for predicting the service life of the optical fiber and helps identify sections requiring replacement or repair. Finally, by detecting the brittleness of the optical fiber structure based on the fiber fatigue life data, structural damage incurred under fatigue loads can be promptly detected. This series of analysis steps provides a more comprehensive and accurate assessment of the health of the optical cable, avoiding the limitations of traditional methods that rely solely on image recognition. By integrating the physical and chemical properties of the optical cable, it provides a more accurate assessment of fatigue and corrosion damage, improving the accuracy, efficiency, and long-term stability of optical cable inspection. These beneficial effects will greatly improve the feasibility and reliability of the optical cable monitoring system in practical applications, while reducing the failure rate and extending the service life of the optical cable.

[0042] Preferably, the step S2 of enhancing the structural stability of the exposed section optical cable data includes:

[0043] Statistics of highly brittle structure optical fiber data are collected based on optical fiber structure brittle data;

[0044] Determine the brittle cable section based on the exposed section optical cable data based on the highly brittle structure optical fiber data;

[0045] Designing UV-resistant coating based on brittle cable segments;

[0046] Laying corrosion-resistant connectors based on the brittle optical cable segment; enhancing the sealing performance of the corrosion-resistant connectors; performing laser surface treatment on the brittle optical cable segment based on the sealing performance to obtain laser surface data; performing micro-nanostructure design on the brittle optical cable segment based on the laser surface data to obtain micro-nanostructure data;

[0047] The structural stability of the exposed section optical cable data is evaluated based on the micro-nano structure data to obtain the structural stability data.

[0048] By collecting statistical data on highly brittle structural optical fibers, the present invention can identify the most severely damaged optical fiber areas, providing clear targets for subsequent reinforcement and repair measures. Based on these highly brittle optical fiber segments, the brittle areas in the exposed sections of the optical cable can be accurately divided, thereby providing data support for the design of targeted protection measures for this section of the optical cable. Designing and laying anti-ultraviolet coatings can effectively delay the damage of ultraviolet rays to the optical fiber, reduce the degradation of optical fiber performance caused by ultraviolet radiation, and improve the weather resistance and service life of the optical fiber. At the same time, selecting suitable anti-corrosion connectors and enhancing their sealing can significantly reduce the damage to the optical fiber caused by environmental moisture, corrosive gases, etc., and further improve the stability of the optical cable in harsh environments. By enhancing the sealing and combining it with laser surface treatment, not only can surface contamination and tiny cracks be removed, but the optical fiber surface can also be further optimized to improve the durability of the optical fiber. Based on laser surface treatment data, micro-nano structure design helps to improve the damage resistance and structural strength of the optical fiber, ensuring the stability and reliability of the optical fiber after long-term use. This series of steps, through the comprehensive use of surface treatment, coating design, connector optimization and other technical means, can effectively improve the durability and damage resistance of brittle optical cable segments from multiple aspects, thereby enhancing the overall structural stability of the optical cable, ensuring its long-term reliable operation in harsh environments, and providing innovative solutions for optical cable maintenance and service life extension.

[0049] Preferably, step S3 is specifically as follows:

[0050] Step S31: constructing an optical cable defect image recognition model based on optical fiber structure embrittlement data;

[0051] Step S32: calculating the folding angle of the folded section optical cable data according to the optical cable defect image recognition model;

[0052] Step S33: locating abnormal points of microbending loss based on the folding angle;

[0053] Step S34: Calculating the loss according to the microbending loss abnormal point to obtain microbending loss data;

[0054] Step S35: locating the break point of the folded section optical cable data based on the microbending loss data;

[0055] Step S36: Identify the optical fiber cross-sectional structural features based on the breakpoint.

[0056] The present invention uses an optical cable defect image recognition model constructed based on fiber structural embrittlement data to automatically identify potential defects and damaged areas in optical cables, reducing reliance on manually annotated data and improving the model's accuracy and robustness. By calculating the fold angle of a folded section of optical cable, the specific location of cable damage, particularly minor fold damage, can be precisely located, providing a basis for further analysis of microbend loss. Locating microbend loss anomalies helps accurately identify the loss area of the optical cable and assess the degree of performance degradation, thereby improving the efficiency and accuracy of damage identification. The calculated microbend loss data provides clear data support for subsequent fault location, making the repair or replacement of the damaged cable more targeted. Based on the microbend loss data, the break point is located, further reducing the damaged area, providing high-precision positioning for optical cable repair, ensuring timely repair and preventing further damage caused by cable breakage. Ultimately, by identifying the structural characteristics of the optical fiber cross section, a deeper understanding of the causes and effects of optical fiber damage can be achieved, facilitating optimization during the cable design phase and improving the structural stability and durability of the cable. Through the synergistic effect of this series of steps, the efficiency of optical cable defect detection can be effectively improved, the optical cable maintenance process can be optimized, the probability of false detection and missed detection can be reduced, and the stability and long-term performance of optical cables in complex environments can be improved.

[0057] Preferably, step S35 is specifically as follows:

[0058] Step S351: performing fiber Bragg grating measurement on the folded section optical cable data to obtain grating measurement data;

[0059] Step S352: Calculating strain based on the grating measurement data to obtain strain data;

[0060] Step S353: drawing a strain curve of the optical fiber according to the strain data; identifying a region where strain increases suddenly based on the strain curve of the optical fiber; and identifying a region where strain decreases suddenly based on the strain curve of the optical fiber.

[0061] Step S354: performing a region intersection operation based on the strain sudden increase region and the strain sudden decrease region to obtain a suspected fracture point region;

[0062] Step S355: verifying the suspected fracture point region based on the microbending loss data to obtain the fracture point region;

[0063] Step S356: Locate the break point of the folded segment optical cable data according to the break point region.

[0064] The present invention can obtain accurate grating measurement data by performing fiber Bragg grating measurement on the folded section optical cable data, providing a reliable basis for subsequent strain analysis. The strain data calculated based on the grating measurement data can effectively reflect the deformation of the optical fiber under different stress states, and provide key data for judging whether the optical fiber is deformed or damaged. The optical fiber strain curve is drawn according to the strain data, and the strain sudden increase area and strain sudden drop area can be further identified through the curve, which helps to find the local damage point of the optical fiber, especially the risk area of stress concentration or material fracture. By performing an intersection operation on the strain sudden increase and drop areas, the suspected break point area can be accurately identified, providing a basis for the precise positioning of the break point. Combining the verification of the suspected break point area with microbend loss data can effectively improve the accuracy of break point identification and ensure the authenticity and effectiveness of the break point area. Finally, by locating the break point area, the break point of the folded section optical cable can be accurately determined, providing reliable technical support for the repair, maintenance or replacement of the optical cable, and avoiding potential risks caused by missed detection or false detection. Through this series of steps, the accuracy, efficiency and reliability of optical cable defect detection can be significantly improved, the impact of optical cable failures on communication systems can be reduced, and the service life and stability of optical cables can be improved.

[0065] Preferably, step S36 is specifically as follows:

[0066] Step S361: Detecting brittle fracture based on the fracture point to obtain brittle fracture data; detecting ductile fracture based on the fracture point to obtain ductile fracture data; detecting microcrack propagation fracture based on the fracture point to obtain microcrack propagation fracture data;

[0067] Step S362: Identifying radial cracks based on brittle fracture data;

[0068] Step S363: detecting fibrous plastic deformation based on ductile fracture data;

[0069] Step S364: detecting fracture surface roughness based on microcrack propagation fracture data;

[0070] Step S365: Integrate radial cracks, fiber-shaped plastic deformation, and fracture roughness to obtain fiber cross-sectional structural characteristics.

[0071] By detecting brittle fracture, ductile fracture, and microcrack propagation fracture based on the fracture point, the present invention can comprehensively assess the fracture type of the optical cable, providing detailed data for subsequent fracture analysis. Brittle fracture data helps identify the fragility of optical fiber materials under external forces and detect the risk of optical fiber fracture, while ductile fracture data reveals the fracture behavior of optical fibers under large deformations, providing a basis for analyzing the toughness of the optical fiber. Microcrack propagation fracture data can help identify the propagation path of microcracks before fracture, further revealing vulnerable areas of the optical fiber structure and providing an important reference for preventing small cracks from triggering more serious fractures. Identifying radial cracks based on brittle fracture data helps locate cracks on the surface or inside the optical fiber, which are the root cause of optical cable damage. Detecting fiber-shaped plastic deformation based on ductile fracture data reveals the plastic deformation behavior of the optical fiber under external forces, helping to assess whether the optical fiber can withstand external pressure for a long time without failure. By detecting fracture surface roughness based on microcrack propagation fracture data, the morphological characteristics of the fracture site can be identified, providing a basis for further evaluating the durability of the optical fiber. By integrating radial cracks, fiber plastic deformation, and fracture roughness data, we can fully understand the cross-sectional structural characteristics of the optical fiber, thereby providing a reliable basis for optical fiber health status assessment and maintenance decisions, ensuring the long-term stable operation of the optical cable and reducing the occurrence of failures.

[0072] Preferably, this specification also provides an optical cable defect detection system based on an image recognition model, which is used to execute the optical cable defect detection method based on an image recognition model as described above. The optical cable defect detection system based on an image recognition model includes:

[0073] The optical cable laying morphology recognition module is used to obtain an optical cable laying image; identify the optical cable laying morphology based on the optical cable laying image, and obtain the exposed section optical cable data and the folded section optical cable data;

[0074] The structural stability enhancement module is used to evaluate the corrosion resistance of the optical cable based on the exposed section optical cable data; determine the degree of armor rust based on the optical cable corrosion resistance; detect the brittleness of the optical fiber structure based on the degree of armor rust and obtain optical fiber structural brittleness data; enhance the structural stability of the exposed section optical cable data based on the optical fiber structural brittleness data and obtain structural stability data;

[0075] The optical fiber cross-sectional structure recognition module is used to construct an optical cable defect image recognition model based on optical fiber structural embrittlement data; detect microbend loss in the folded section optical cable data according to the optical cable defect image recognition model to obtain microbend loss data; locate the breakpoint of the folded section optical cable data based on the microbend loss data; and identify the optical fiber cross-sectional structure characteristics based on the breakpoint;

[0076] The optical cable structure defect detection module is used to transmit the optical fiber cross-sectional structural characteristics and structural stability data to the optical cable defect image recognition model, detect the optical cable structure defects in the optical cable laying image, and generate optical cable structure defect data.

[0077] The optical cable defect detection system based on an image recognition model of the present invention can implement any optical cable defect detection method based on an image recognition model of the present invention, and is used to combine the operation and signal transmission medium between various modules to complete the optical cable defect detection method based on the image recognition model. The internal modules of the system cooperate with each other to enhance the accuracy and robustness of optical cable defect detection, thereby improving the accuracy of optical cable defect identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0079] Figure 1 This is a schematic flow chart of the steps of an optical cable defect detection method based on an image recognition model according to the present invention;

[0080] Figure 2 Detailed step flow diagram of step S1 in the present invention;

[0081] Figure 3 Detailed step flow diagram of step S3 in the present invention;

[0082] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0083] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0084] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0085] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0086] To achieve this, please refer to Figures 1 to 3 The present invention provides a method for detecting optical cable defects based on an image recognition model, the method comprising the following steps:

[0087] Step S1: Acquire an optical cable laying image; identify the optical cable laying form according to the optical cable laying image, and obtain exposed section optical cable data and folded section optical cable data;

[0088] In this embodiment, an image of the optical cable laying is obtained, and an HD camera or drone is used to collect images of the optical cable laying site, ensuring that the image resolution is not less than 4K so that the shape of the optical cable can be clearly identified. Image preprocessing technology is used to perform noise reduction processing to filter out illumination changes and noise interference in the image. An image segmentation algorithm, such as the Canny algorithm based on edge detection, is used to segment the image to distinguish between exposed sections of optical cable and folded sections of optical cable. The exposed section of optical cable is defined as the part that is not covered by other objects and is directly exposed to the outside, while the folded section of optical cable is the part that has been bent or folded to a certain extent. In image processing, by setting a suitable edge detection threshold, usually set to between 40 and 60, it is ensured that the edge of the optical cable in the image is clearly extracted. By analyzing the shape of the optical cable and the laying environment, the exposed section of optical cable data and the folded section of optical cable data are obtained. These data include geometric features such as the length, curvature and bending radius of the optical cable.

[0089] Step S2: Evaluating the corrosion resistance of the optical cable based on the exposed section optical cable data; determining the degree of armor rust based on the optical cable corrosion resistance; detecting the embrittlement of the optical fiber structure based on the degree of armor rust to obtain optical fiber structure embrittlement data; enhancing the structural stability of the exposed section optical cable data based on the optical fiber structure embrittlement data to obtain structural stability data;

[0090] In this embodiment, based on the data of the exposed section of optical cable, a fiber corrosion resistance assessment method is used to obtain the degree of corrosion of the optical cable surface or armor layer through sensors or electrochemical corrosion testing methods. Specifically, an electrochemical tester is used to measure the conductivity of the armor layer material to obtain the degree of rust of the armor layer. The value range of the rust degree is 0 to 100, where 0 means no rust and 100 means complete rust. Then, according to the degree of rust of the armor layer, a threshold value is set (for example, when the rust degree is 50) to detect the embrittlement of the optical fiber structure. In this process, a strain tester is used to measure the strain value of the exposed section of optical cable, and based on these strain data, it is determined whether the optical fiber structure is affected by corrosion. The embrittlement data of the optical fiber structure is obtained by comparing the strain values at different positions. When the strain value exceeds the set threshold (such as more than 0.05%), it is marked as an embrittled area. By enhancing the optical fiber structure data of the embrittled area, the structural stability of the exposed section of optical cable data is increased. Using a numerical simulation method, the stress distribution of the embrittled area is calculated through finite element analysis to enhance its structural stability.

[0091] Step S3: constructing an optical cable defect image recognition model based on the optical fiber structure embrittlement data; detecting microbending loss of the folded section optical cable data according to the optical cable defect image recognition model to obtain microbending loss data; locating the breakpoint of the folded section optical cable data based on the microbending loss data; and identifying optical fiber cross-sectional structural features based on the breakpoint.

[0092] In this embodiment, a diverse dataset of optical cable images, including defect images of exposed and folded sections, was collected. Data augmentation methods, such as rotation, scaling, and translation, were used to expand the diversity of the training dataset. A convolutional neural network (CNN) was used for training, and model parameters were optimized using a backpropagation algorithm. During model training, cross entropy was used as the loss function, the number of training iterations was set to 1000, and the learning rate was 0.001. During training, the network hyperparameters were adjusted based on the performance of the validation set. After model training, the model was used to detect microbend loss in folded sections of optical cable. A loss threshold (e.g., a loss value exceeding 0.3dB) was set to determine whether microbend loss was present. By calculating the microbend loss, the break point of the folded section was located. Combined with the geometric morphological characteristics of the cable break, the fiber cross-sectional structural features were further identified. Fiber cross-sectional structural features include cross-sectional morphology, crack direction, and crack width.

[0093] Step S4: Transmitting the optical fiber cross-sectional structural features and structural stability data to the optical cable defect image recognition model, detecting the optical cable structural defects in the optical cable laying image, and generating optical cable structural defect data.

[0094] In this embodiment, optical fiber cross-sectional structural features and structural stability data are transmitted to an optical cable defect image recognition model to detect structural defects in cable installation images. The optical cable structural stability data is used as one of the input conditions to assist the image recognition model in detecting defects in the cable installation images. This process can be divided into the following key steps: data collection and preprocessing: image data is acquired from the cable installation site and preprocessed in conjunction with optical fiber structural embrittlement data. Demarcation of embrittled cable segments provides detailed reference information for subsequent image recognition. Identification of embrittled cable segments: structural stability data, particularly micro- and nanostructure data, is used to train the image recognition model to identify the locations of embrittled cable segments. These locations are often the most prone to defects such as cracks, breakage, and corrosion. Structural stability assessment: Based on the micro- and nanostructure data, the structural stability of the optical cable is assessed. Combined with the defect areas in the image, it is determined whether these defects affect the long-term stability and service life of the optical cable. The optical cable geometric data is processed to extract key features such as bend angle and bend radius as one of the input data. The bend angle can be calculated by calculating the curvature change of the optical cable path, while the bend radius is derived using the formula for calculating the minimum bend radius of the optical cable. Next, combining the fiber's breakpoint and microbend loss data, features related to cable structural defects are extracted. Microbend loss is determined by analyzing the fiber's loss in the folded section. Typically, a loss value greater than 0.3dB indicates the presence of microbend loss in the cable, serving as a defect indicator. The data transmitted to the model includes fiber cross-sectional structural features, such as the breakpoint (located from the previous step), crack characteristics (crack width and direction extracted through image analysis), and structural stability data, such as the brittle area (derived from the brittleness analysis) and corrosion resistance level (derived from corrosion severity assessment). After data formatting and standardization, this data is fed into the optical cable defect image recognition model. The model utilizes a convolutional neural network (CNN) architecture trained on a large number of optical cable defect images and associated structural feature data. The model analyzes the input image data and its associated structural features, and determines whether the cable structure is defective using a set threshold. When the crack width exceeds 0.2mm or the microbend loss at the breakpoint exceeds 0.3dB, the model identifies a structural defect in the cable. The model generates cable structural defect data detailing the location, type, and severity of the detected defect, along with recommended repairs. These results serve as a basis for subsequent cable maintenance, repair, and monitoring, helping personnel promptly identify potential risks and conduct necessary inspections and maintenance.

[0095] Preferably, step S1 is specifically as follows:

[0096] Step S11: Acquire an optical cable laying image;

[0097] In this embodiment, images of the optical cable installation are captured using a high-resolution camera to ensure clarity and fully demonstrate the cable surface features. The camera resolution must be at least 300 dpi to ensure the integrity of image detail. During capture, the image exposure time must be moderate to avoid overexposure or underexposure, which can affect the capture of cable surface details. The shooting angle must take into account the overall cable installation direction and specific high-risk areas. The cable installation image should be free of obstructions to ensure consistent image quality.

[0098] Step S12: removing low-frequency noise from the optical cable laying image to obtain a noise-reduced optical cable laying image;

[0099] In this embodiment, low-frequency noise in the cable laying image is removed using a two-dimensional high-pass filter. A standard high-pass filter, such as a 2D Gaussian filter, is selected. The filter cutoff frequency is set to 0.1 Hz and adjusted based on the frequency domain characteristics of the image. The cable laying image is converted to the frequency domain, and a Fourier transform is performed to obtain a frequency spectrum. The low-frequency portion is then removed using a high-pass filter. The filtered image is then returned to the spatial domain using an inverse Fourier transform. This operation effectively removes low-frequency noise from the image, ensuring a clearer image for subsequent processing.

[0100] Step S13: Detecting the edge of the optical cable based on the denoised image of the optical cable laying, wherein the upper and lower detection thresholds are set to 50 and 150, respectively, to obtain the edge of the optical cable, and extracting the candidate optical cable area of the denoised image of the optical cable laying according to the edge of the optical cable;

[0101] In this embodiment, the edge of the optical cable is detected based on the noise reduction image of the optical cable laying, and the Canny edge detection algorithm is used for edge detection. First, the high and low thresholds are set to 50 and 150, and the edges in the image are calculated by gradient. The image is first smoothed, and the noise is reduced by a Gaussian filter. Then the gradient value of the image is calculated to obtain the edge candidate points. Then, according to the set high and low thresholds, the area with a gradient value greater than 150 is marked as an edge, and the area with a gradient value less than 50 is excluded. The obtained optical cable edge will be used to extract the optical cable candidate area of the optical cable laying noise reduction image. The boundary of the optical cable candidate area is obtained by connecting the edge points to ensure that the candidate area accurately surrounds the main part of the optical cable.

[0102] Step S14: performing an expansion operation on the optical cable candidate region using a preset 3x3 structure element to obtain an optical cable expansion region;

[0103] In this embodiment, a dilation operation is performed on the candidate optical cable region using a preset 3x3 structuring element. This dilation operation utilizes a classic morphological dilation algorithm to expand the boundaries of the candidate optical cable region to capture edge details. During this process, a 3x3 structuring element is selected. During the dilation operation, each pixel in the candidate region is updated based on the maximum value of its neighboring pixels. The dilation operation is implemented by comparing the value of each pixel in the candidate optical cable region with the values of other pixels within the area covered by the structuring element. The maximum pixel value is taken as the dilated region value, ultimately resulting in the dilated optical cable region.

[0104] Step S15: performing an erosion operation on the optical cable candidate area using a preset 5x5 structure element to obtain an optical cable erosion area;

[0105] In this embodiment, an erosion operation is performed on the candidate optical cable regions using a preset 5x5 structuring element. This operation is implemented using a morphological erosion algorithm. The goal is to reduce the boundaries of the candidate optical cable regions and remove small areas of the optical cable. During the erosion process, using a 5x5 structuring element, the pixel value of each candidate region is compared with the minimum pixel value in the neighborhood covered by the structuring element. The minimum value is then selected to update the center pixel. This operation effectively eliminates some noise points and makes the boundaries of the candidate regions clearer.

[0106] Step S16: Calculate the aspect ratio of the cable corrosion area. If the aspect ratio is greater than 2:1, obtain the exposed cable segment data. Calculate the curvature of the cable corrosion area. If the curvature radius is less than 10 cm, obtain the folded cable segment data.

[0107] In this embodiment, the aspect ratio of the cable corrosion area is calculated. The aspect ratio is obtained by measuring the minimum circumscribed rectangle of the cable corrosion area. First, the contour of the cable corrosion area is extracted to determine the length and width of the minimum circumscribed rectangle. Then, by calculating the aspect ratio, if the aspect ratio is greater than 2:1, the area is judged to be an exposed section of optical cable. The exposed section of optical cable usually appears longer and narrower. Next, the curvature of the cable corrosion area is calculated. The contour of the corrosion area is fitted using a curve fitting algorithm to obtain the curvature of the curve. If the curvature radius is less than 10 cm, the area is judged to be a folded section of optical cable. The folded section of optical cable usually has a smaller bending radius. By determining the aspect ratio and curvature value, the exposed section optical cable data and the folded section optical cable data are finally obtained, providing data support for subsequent optical cable defect detection.

[0108] Preferably, the evaluating the corrosion resistance of the optical cable in step S2 includes:

[0109] Extracting surface texture data of exposed optical cable segment data;

[0110] In this embodiment, images of the exposed section of optical cable are captured using a high-precision camera with a resolution of at least 300 dpi to obtain images of the cable surface. Image processing uses grayscale technology to convert the image into a single-channel grayscale image to facilitate the subsequent extraction of texture features. Subsequently, a Gabor filter is used to perform texture analysis on the cable surface image. The Gabor filter parameters are set to a wavelength of 3 pixels, and the directional angles are selected as 0°, 45°, 90°, and 135° to cover texture information in different directions. The filtered image is used to extract local texture features of the cable surface, obtaining texture data reflecting the cable surface details, including information on the cable surface roughness, crack characteristics, and other surface irregularities.

[0111] Calculating gradients based on surface texture data;

[0112] In this embodiment, the Sobel operator is used to calculate the image gradient on the acquired surface texture data. The convolution kernels of the Sobel operator are [-1, 0, 1] and [-1, 2, -1], respectively. Convolution operations are performed on the image in the horizontal and vertical directions to obtain the gradient value of the image. The gradient value reflects the magnitude of the change in the surface of the optical cable and is used to characterize the intensity of the texture information. During the implementation process, the surface texture data is first smoothed and denoised using a 3x3 Gaussian filter to prevent noise from affecting the gradient calculation. The calculated gradient value is stored at each pixel position in the image to obtain the gradient magnitude and direction of each point, providing basic data for the subsequent calculation of the local gradient change rate.

[0113] Calculate the local gradient change rate based on the gradient;

[0114] In this embodiment, the local gradient change rate calculation uses the gradient change in the local area to measure the subtle changes in the surface texture of the optical cable. In specific implementation, a 3x3 local window is selected and the local gradient difference is calculated around each pixel position. The local gradient change rate is calculated by differentiating the gradient values of each pixel in the window. The differential calculation formula is: local gradient change rate = | gradient (x, y) - gradient (x+1, y) | + | gradient (x, y) - gradient (x, y+1) |, where (x, y) is the pixel position at the center of the window. If the local gradient change rate exceeds the set threshold (such as 0.2), it is considered that there is a significant surface change at that location, which is a corrosion or other defect area, providing a basis for the subsequent detection of corrosion areas.

[0115] Determine the optical cable corrosion area of the optical cable laying image based on the local gradient change rate;

[0116] In this embodiment, when determining the optical cable corrosion area based on the local gradient change rate, a threshold is first set (for example, a gradient change rate greater than 0.2). By comparing it with the calculated local gradient change rate, areas with a larger gradient change rate are screened out. Then, combined with other structural features in the optical cable image (such as edge detection results), a region growing algorithm is used to further extract areas with significant gradient changes. These areas are typically where corrosion or damage exists on the optical cable surface, and specific morphological operations (such as dilation and corrosion) can be used to extract a more accurate outline of the corrosion area. Ultimately, the optical cable corrosion area obtained through these steps is used for subsequent depth and rate calculations.

[0117] Identify the corrosion depth based on the corroded area of the optical cable;

[0118] In this embodiment, the depth of optical cable corrosion is identified using three-dimensional image reconstruction technology. First, representative corrosion samples are selected from different areas of the optical cable. High-definition cameras are used to capture multi-angle images of these areas, and laser scanning technology is used to obtain microscopic depth information of the cable surface. Based on this data, image processing algorithms are used to extract the surface contours, and combined with a least-squares surface fitting technique, the three-dimensional depth of the corroded area is determined. A corrosion depth threshold is set (e.g., greater than 1 mm). If the corrosion depth exceeds this threshold, the area is marked as severely corroded.

[0119] Calculate the corrosion rate based on the corrosion area of the optical cable;

[0120] In this embodiment, the calculation of the optical cable corrosion rate is performed by time series image analysis. When the optical cable laying image is first acquired, the starting depth of each optical cable corrosion area is recorded, and then the optical cable image is acquired again within a certain time interval, and the latest corrosion depth data is obtained through the same three-dimensional reconstruction and depth extraction process. Based on these depth change data, the corrosion rate is calculated. The corrosion rate formula is: corrosion rate = (final depth - initial depth) / time interval. In the specific values, the initial depth and final depth are taken from image data at different time points, and the time interval is the difference between the shooting time and the last shooting time. Through multiple data collection and comparison, the corrosion rate value of each corrosion area is obtained.

[0121] The corrosion resistance of the exposed section of optical cable is evaluated based on the corrosion depth and corrosion rate.

[0122] In this embodiment, a specific corrosion resistance evaluation standard is set. Areas with a corrosion depth exceeding 2mm and a corrosion rate exceeding 0.1mm / year are defined as having poor corrosion resistance, which will affect the normal use of the optical cable and therefore need to be repaired in a timely manner. The values of the corrosion depth and corrosion rate are obtained through the aforementioned steps and recorded in combination with the specific location and time of the corrosion area of the optical cable. During implementation, each corroded area of the optical cable is analyzed one by one, and the initial corrosion depth and corrosion rate data of each area over time are first obtained. Then, by comparing these data with the set standards, a classification judgment is made. If the corrosion depth is greater than 2mm and the corrosion rate exceeds 0.1mm / year, the area is marked as "poor corrosion resistance" and a maintenance alarm is issued based on the result; if the corrosion depth and rate do not reach the set threshold, the corrosion resistance of the area is considered to be good and can still be monitored. All areas marked as poor corrosion resistance will be repaired first, while areas with good corrosion resistance will be included in the scope of regular inspection and monitoring to ensure long-term stable operation.

[0123] Preferably, determining the degree of rust of the armor layer in step S2 includes:

[0124] Corrosion simulation was performed based on the corrosion resistance of the optical cable, wherein the temperature range was set to 25°C-40°C and the humidity range was set to 30%RH-98%RH to obtain corrosion simulation data;

[0125] In this embodiment, the simulation of the corrosion resistance of the optical cable first requires setting the environmental conditions, specifically the temperature range is 25°C to 40°C, and the humidity range is 30% RH to 98% RH. Use environmental simulation equipment (such as a constant temperature and humidity chamber) to expose the optical cable samples to ensure that they operate within the set range. The temperature and humidity in the equipment are continuously monitored and adjusted in real time by built-in sensors to ensure that the temperature and humidity conditions do not deviate from the set range. The temperature and humidity data need to be recorded once an hour for subsequent corrosion analysis. Typically, the simulation time is 48 to 72 hours to ensure that the optical cable is fully corroded under these environments.

[0126] Calculate the thickness of the optical cable armor layer based on corrosion simulation data;

[0127] In this embodiment, after the corrosion simulation is completed, a precision laser thickness gauge (such as the TR800 model) is used to measure the optical cable armor layer point by point to ensure high-precision thickness data. The data at each measurement point should be recorded in a digital database, and the measurement accuracy is required to reach the micron level. For example, measurements are taken at every 10 cm interval to obtain the thickness data of the optical cable armor layer. The measurement results include the actual armor layer thickness at each point, and data smoothing is usually required before analysis to eliminate external interference. Each measured value is compared with a reference thickness (such as 0.8 mm) to ensure the reliability of the results.

[0128] Calculate and obtain thickness reduction data using the preset optical cable armor layer benchmark thickness and the optical cable armor layer thickness;

[0129] In this embodiment, the thickness reduction of each area is calculated based on the thickness data of the optical cable armor layer. First, a reference thickness value is set, such as the normal thickness of the optical cable armor layer is 0.8mm. For each area, the actual thickness value measured (for example, 0.7mm) is subtracted from the reference thickness to obtain the thickness reduction value of the area (for example, 0.8mm-0.7mm=0.1mm). Through this method, each measurement point is analyzed to obtain the thinning distribution data of the optical cable armor layer. These thinning values are used for subsequent corrosion depth and armor rust degree assessment.

[0130] The electrical conductivity of the corrosion simulation data was recorded using an eddy current detector;

[0131] In this embodiment, an eddy current detector (e.g., ET-3000) is used to detect the optical cable armor layer to obtain conductivity data under corrosion simulation conditions. The eddy current detector reflects the conductivity characteristics of the optical cable armor layer by generating an alternating current and monitoring the changes in the current in the conductive material. The device obtains conductivity data related to the degree of corrosion of the optical cable armor layer by reading the signal changes. Tests are performed at each point and the conductivity values are recorded. The conductivity of each corroded area is compared with the conductivity value of the uncorroded area (e.g., normally 0.9S / m) for subsequent analysis of the corrosion depth.

[0132] Determine corrosion depth based on electrical conductivity;

[0133] In this example, an established empirical model is used to analyze the conductivity data recorded by an eddy current detector to determine the corrosion depth of the optical cable armor layer. The relationship between conductivity and corrosion depth is obtained from experimental data and converted using a mathematical formula. For example, an empirical formula is set: a conductivity of 0.9S / m indicates no corrosion, and a conductivity below 0.3S / m corresponds to a corrosion depth of 1.5mm. Using this formula, the corrosion depth can be calculated based on the conductivity value of each area and recorded in a database. This data will be used to subsequently evaluate the degree of armor rust.

[0134] The degree of rust of the armor layer is determined based on the corrosion depth and thickness reduction data.

[0135] In this embodiment, the specific degree of armor damage is determined based on the relationship between corrosion depth data and thickness reduction data. For example, specific criteria for corrosion depth and thickness reduction are set: when the corrosion depth is greater than 1.0mm and the thickness reduction exceeds 20%, the armor is classified as "severely rusted"; when the corrosion depth is between 0.5mm and 1.0mm and the thickness reduction is between 10% and 20%, it is classified as "moderately rusted"; and when the corrosion depth is less than 0.5mm and the thickness reduction is less than 10%, it is classified as "slightly rusted." Each corroded area is classified based on its corrosion depth and thickness reduction values, combined with these thresholds, to determine its specific rust level. During this process, corrosion depth and thickness reduction data are calculated for each corroded area of the optical cable. A pre-defined formula is used to correlate the corrosion depth of each area with the degree of armor thickness reduction, thereby determining the rust level for each area. If an area is assessed as severely rusted, it is marked as a priority area and recommended for repair or replacement. The rust classification results will be recorded and serve as an important basis for subsequent optical cable maintenance decisions.

[0136] Preferably, the detecting of optical fiber structure embrittlement in step S2 includes:

[0137] Determine the rust area according to the rust degree of the armor layer;

[0138] In this embodiment, the rusted areas of the optical cable armor layer are calibrated one by one based on the corrosion depth and armor layer thickness thinning data. Standard values for corrosion depth and thickness thinning are set. For example, if the corrosion depth exceeds 0.5 mm or the thickness thinning exceeds 10%, the area can be regarded as a rusted area. The conductivity data obtained by the eddy current detector is combined with the known corrosion simulation data to perform detailed measurements on each optical cable segment. The corrosion depth and thickness thinning values of each optical cable segment are compared with the preset rust classification standards. If a certain rust grade standard (such as light rust, heavy rust) is met, it is calibrated as the corresponding rust area. For example, an area with a corrosion depth greater than 1.0 mm and a thinning of more than 20% is calibrated as heavy rust. Data records will be classified in the system to ensure accurate calibration of rusted areas.

[0139] Determine the degree of armor failure based on the rusted area;

[0140] In this embodiment, each optical cable segment marked as a rusty area is further analyzed, and the degree of failure of the area is determined in combination with the structural characteristics of the armor layer. Specific standard values are set. For example, if the corrosion depth is greater than 0.5 mm and the thinning is greater than 10%, the area is determined to be a mild failure; if the corrosion depth exceeds 1.0 mm and the thinning exceeds 20%, it is a severe failure. The degree of failure is determined using the material properties of the armor layer and data such as corrosion depth and thickness thinning. This process can further refine the determination of the degree of failure using corrosion simulation environment data (such as temperature and humidity). Through these steps, the degree of failure of each optical cable segment is classified and recorded, and data support is provided for subsequent optical cable maintenance and repair decisions.

[0141] Divide the exposed cable data into armored rusted cable segments based on the degree of armor failure;

[0142] In this embodiment, the exposed sections of the optical cable are divided according to the degree of failure of the armor layer. The optical cable sections with a higher degree of failure are divided into "armor layer rusted optical cable sections". For example, the areas where the armor layer has moderate or severe failure are divided into rusted optical cable sections. By combining the working environment of the optical cable (such as buried or overhead) and actual usage, the failure degree of the exposed sections of the optical cable is further analyzed, and the areas are divided according to the preset failure standards (such as corrosion depth greater than 0.5mm and thickness thinning greater than 10%). All division results will be entered into the database for storage, and a failure area map will be generated to provide detailed information for subsequent optical cable maintenance and optimization.

[0143] The fiber bending load is simulated for the rusted armored cable section, and the fiber fatigue life is predicted to obtain the fiber fatigue life;

[0144] In this embodiment, the optical fiber bending load simulation is performed on the rusty optical cable section of the armor layer to simulate the bending load that the optical fiber is subjected to in actual use. First, specific bending load parameters (such as bending radius, bending angle, maximum load value of optical fiber, etc.) are set. Then, simulation calculations are performed based on finite element analysis (FEA) to obtain the stress distribution and fatigue damage of the optical fiber under different load conditions. According to the material, bending angle and stress conditions of the optical fiber, the fatigue life of the optical fiber is predicted. For example, when the bending radius is 50mm and the maximum load of the optical fiber is 100N, the fatigue life of the optical fiber can be calculated by simulation. Appropriate material parameters (such as elastic modulus, yield strength, etc. of the optical fiber) need to be set during the simulation process to ensure the accuracy of the results. Finally, the fatigue life data of each optical fiber is obtained through bending load simulation.

[0145] The optical fiber structure embrittlement is detected based on the optical fiber fatigue life to obtain the optical fiber structure embrittlement data.

[0146] In this embodiment, non-destructive testing methods, such as fiber optic reflectometry (FOT) and optical time-domain reflectometry (OTDR), are used to monitor optical fiber microdamage, cracks, and surface hardening. OTDR technology, by emitting light pulses and analyzing the returned signals, can accurately detect damage locations, cracks, and other structural defects within the optical fiber, thereby assessing the fiber's integrity. Furthermore, fiber optic reflectometry can help identify the presence of surface hardening or microcracks by measuring changes in the reflected signal intensity. A criterion for determining fiber structural embrittlement is established in conjunction with fiber fatigue life prediction data. For example, if the fatigue life of an optical fiber is less than 1000 bending cycles, it indicates a risk of embrittlement after long-term load exposure. This threshold can be adjusted based on actual application requirements to ensure early detection of embrittlement. Next, embrittlement is determined by measuring the attenuation of the optical fiber transmission signal. Specifically, a fiber attenuation tester is used to assess the attenuation characteristics of the optical fiber. When the attenuation exceeds a set threshold (e.g., 0.5 dB / km), it indicates that the optical fiber's transmission capacity has decreased, indicating the presence of surface damage or internal defects. In addition to attenuation, detecting cracks on the optical fiber surface is also crucial. Crack detection can be performed using a microscope or infrared scanning to check whether microcracks appear on the surface of the optical fiber. If the crack length exceeds the standard threshold (such as 0.2mm), it is judged that the optical fiber has become brittle. In addition, strain data can be used to monitor whether the optical fiber is overstretched or bent. If the strain value exceeds the preset threshold (such as 200μm / m), it indicates that the optical fiber has structural damage under long-term stress. Through the comprehensive analysis of these data, the degree of brittleness of the optical fiber can be more accurately assessed. Finally, the fatigue life prediction data, attenuation value, crack condition and strain data are integrated to obtain the overall brittleness condition of the optical fiber, and finally generate the optical fiber structural brittleness data. These data provide a scientific basis for subsequent optical cable maintenance and repair, helping to determine which optical fiber segments need to be replaced or reinforced for protection.

[0147] Preferably, the step S2 of enhancing the structural stability of the exposed section optical cable data includes:

[0148] Statistics of highly brittle structure optical fiber data are collected based on optical fiber structure brittle data;

[0149] In this embodiment, all detected optical fiber brittleness data are collected based on the optical fiber structural brittleness detection results. The structural brittleness of each optical fiber will be evaluated based on fatigue life data, attenuation value, crack detection results and strain data. For each optical fiber, a preset threshold value (such as an attenuation value greater than 0.5dB / km, a crack length greater than 0.2mm, a strain value greater than 200μm / m, etc.) is used to determine whether it belongs to a highly brittle structural optical fiber. When any of the above conditions is met, the optical fiber is recorded as a highly brittle optical fiber. During the statistical process, an automated data screening tool is used to batch process all optical fibers using the set threshold value, thereby screening out highly brittle optical fiber segments that meet the standards. This process can be performed through data processing software, using standard statistical analysis methods (such as mean, variance analysis, etc.) to quickly obtain the number and distribution location of highly brittle optical fibers.

[0150] Determine the brittle cable section based on the exposed section optical cable data based on the highly brittle structure optical fiber data;

[0151] In this embodiment, the position and status of each fiber in the exposed section of the optical cable are determined by inspecting each section of the optical fiber. Based on this, the highly brittle fiber area is designated as a brittle cable segment according to the fiber brittleness data and the overall structure of the optical cable. The specific method is to set a brittleness standard for the brittle fiber. For example, when the brittleness of the optical fiber reaches 50% or above, the section in which it is located is determined to be a brittle cable segment. Based on the structure of the optical cable, if the length of the brittle fiber segment exceeds a predetermined threshold (e.g., 10m), the segment is designated as a brittle cable segment. This process can be accomplished through a combination of segmented inspection and data analysis, using the distribution and location of highly brittle fibers in combination with the cable layout to determine the brittle cable segment.

[0152] Designing UV-resistant coating based on brittle cable segments;

[0153] In this embodiment, for the optical cable segments that have been determined to be brittle, it is necessary to design an anti-ultraviolet coating. First, a suitable coating material is selected. Such materials need to have strong anti-ultraviolet performance and weather resistance. According to the use environment of the optical cable, select materials with UV protection levels that meet the standards, such as polyurethane, fluoride coating, etc. The thickness of the coating should be designed according to the diameter of the optical cable, the working environment and the durability requirements. The general coating thickness range is 50μm to 200μm. The design of the coating should take into account the intensity of ultraviolet rays and the exposure time of the optical cable to ensure that the coating does not deteriorate under long-term ultraviolet radiation. In addition, the adhesion of the coating and the coating process also need to be strictly controlled, and automated spraying equipment should be used for uniform coating of the coating. After the coating is completed, the coating stability must be tested using a UV aging tester to ensure that it still maintains a stable protective effect under long-term ultraviolet irradiation.

[0154] Laying corrosion-resistant connectors based on the brittle optical cable segment; enhancing the sealing performance of the corrosion-resistant connectors; performing laser surface treatment on the brittle optical cable segment based on the sealing performance to obtain laser surface data; performing micro-nanostructure design on the brittle optical cable segment based on the laser surface data to obtain micro-nanostructure data;

[0155] In this embodiment, corrosion-resistant connectors are used in optical cable systems, particularly in areas exposed to harsh environments, to connect optical cables. These connectors are made of corrosion-resistant materials, such as stainless steel, galvanized steel, or specially coated alloys. They effectively resist moisture, salt spray, high temperatures, and chemical attack, preventing corrosion or structural damage caused by the external environment. The design of the corrosion-resistant connectors ensures a reliable connection with the optical cable, preventing unstable connections or damage to the cable due to environmental factors. The connectors must withstand the mechanical stress and temperature fluctuations of the optical cable, so their structural design must ensure long-term stability. Furthermore, the design and material selection of corrosion-resistant connectors must comply with relevant international standards, such as ISO 9223, to ensure they remain corrosion-resistant in harsh environments for extended periods, thereby ensuring the safety and reliability of the optical cable system. Corrosion-resistant connectors are designed and installed in brittle sections of optical cable, particularly in areas susceptible to corrosion. They should be made of highly corrosion-resistant materials, such as stainless steel, galvanized steel, or alloys with special coatings. During installation, the connectors must ensure a reliable connection with the optical cable and protect the cable from environmental influences. Connectors must be designed to withstand the mechanical stresses and temperature fluctuations of optical cables, maintaining structural stability. Each connector's design must comply with relevant standards and regulations. For example, connectors must meet ISO 9223 corrosion resistance standards to ensure they are corrosion-resistant in harsh environments such as humidity and salt spray. To ensure effective protection at the connection, further measures are taken based on corrosion-resistant connectors to enhance sealing. This step requires the use of high-temperature and corrosion-resistant sealing materials, such as silicone and fluororubber, for the sealing design. The sealing material must meet the requirements of the optical cable's environment, such as heat resistance, weather resistance, and water resistance. The sealing portion of the connector must undergo rigorous pressure testing to ensure that the sealing meets the required standards. The sealing surface must be seamless and provide sufficient compression to prevent moisture and oxygen ingress, thereby effectively preventing corrosion. After enhancing the sealing, the embrittled optical cable sections undergo laser surface treatment to improve surface durability and corrosion resistance. Laser surface treatment typically utilizes laser cladding or laser polishing techniques, which irradiate the optical cable surface with a high-power laser beam to form a hard coating. This coating effectively enhances the cable's UV resistance, corrosion resistance, and abrasion resistance. After surface treatment, laser surface data acquisition equipment is used to test the cable's surface roughness, thickness, and hardness to ensure the treatment meets design requirements. This surface data will provide foundational data support for subsequent micro- and nanostructure design. Optical cable micro- and nanostructures refer to tiny structures formed on the cable surface using micron- and nanometer-scale processing techniques. These structures are designed to enhance specific cable properties. Micro- and nanostructure design utilizes high-precision micromachining techniques, such as photolithography and electron beam lithography, to create functional, small-scale patterns or textures on the cable surface.The purpose of micro-nanostructuring optical cables is to improve their wear resistance, UV resistance, self-cleaning properties, and surface repellency to moisture and contaminants. For example, by forming tiny structures on the surface, the surface roughness can be increased, thereby enhancing the cable's resistance to foreign substances and preventing damage and aging during long-term use. The design of micro-nanostructures takes into account the cable's operating environment, such as high humidity and UV exposure. Therefore, appropriate materials and manufacturing processes are selected during the design process to enhance the cable's adaptability to specific environments. Precision measurement tools such as laser interferometers can be used to obtain data on the size, shape, and distribution of micro-nanostructures to ensure that the design meets the expected functional requirements. Based on this data after laser surface treatment, the micro-nanostructure of the optical cable segment is designed. Micro-nanostructure design utilizes advanced micromachining technologies, using high-precision photolithography and electron beam lithography to create micro-nanostructures with specific functions on the cable surface. This design aims to improve the cable's wear resistance, UV resistance, and surface self-cleaning properties. The designed micro-nanostructures should take into account the cable's operating environment, such as enhancing its repellency to moisture and contaminants. Surface data of micro-nano structures, such as the size, shape and distribution of the microstructures, is obtained through equipment such as laser interferometers to ensure that the design meets the requirements.

[0156] The structural stability of the exposed section optical cable data is evaluated based on the micro-nano structure data to obtain the structural stability data.

[0157] In this embodiment, the structural stability of the exposed cable segment is evaluated based on micro-nanostructure data. First, key data on the micro-nanostructure, such as its size, shape, distribution, surface hardness, and strength, must be collected. Then, a mechanical simulation of the cable segment is performed using finite element analysis (FEA). During the simulation process, the micro-nanostructure data must be input into the model, and appropriate material properties, such as elastic modulus, Poisson's ratio, and yield strength, must be set. Different operating environments (such as temperature changes, humidity, and UV exposure) and stress conditions (such as tension, compression, and bending) for the exposed cable segment must also be simulated. Furthermore, the setting of boundary conditions is also crucial, including the cable's mounting method, load application location, and magnitude. During the simulation analysis, a computer simulates the stress distribution and deformation of the cable under different mechanical loads to assess its compressive, tensile, and bending resistance. Specifically, the computer evaluates whether the cable can maintain sufficient stability under the influence of the micro-nanostructure to prevent structural failure or excessive wear due to improper microstructure design. To ensure the accuracy of the simulation results, the material parameters, boundary conditions, and other simulation settings must be carefully adjusted. By comparing different design schemes, the optimal micro-nanostructure design scheme is selected to ensure that the optical cable can maintain good structural stability in different environments during long-term use and meet various performance requirements in practical applications.

[0158] Preferably, step S3 is specifically as follows:

[0159] Step S31: constructing an optical cable defect image recognition model based on optical fiber structure embrittlement data;

[0160] In this embodiment, optical fiber structure embrittlement data, including physical properties, fatigue life, crack depth and distribution characteristics of optical fibers, are collected, and optical cable image data corresponding to these embrittlement conditions are collected. These images include the appearance of optical cables under different angles and lighting conditions. The images must be clear and marked with different types of defects. Image preprocessing includes denoising, size normalization, and grayscale processing to ensure data quality. Next, an image recognition model based on a convolutional neural network (CNN) is constructed, and the spatial characteristics of defects are automatically learned through image feature extraction modules (such as convolution layers and pooling layers). During the training process, the cross entropy loss function is used for optimization, and the Adam optimizer is used to adjust the learning rate (set to 0.001), batch size (32), and number of iterations (1000 times). The validation set data is used to evaluate the accuracy of the model, and the confusion matrix is used to measure the classification ability of the model. The final model can identify the types of defects in the image, such as cracks, folds, and bends.

[0161] Step S32: calculating the folding angle of the folded section optical cable data according to the optical cable defect image recognition model;

[0162] In this embodiment, a trained defect image recognition model is used to input an optical cable image and identify the folded area in the optical cable. The edges of this area are extracted using an image processing algorithm (such as Canny edge detection), and a straight line fit is performed using the Hough transform to obtain the boundary line of the folded area. Based on the two boundary lines of the folded area, the angle between the two boundary lines is calculated as the fold angle. This angle is calculated using the following formula:

[0163]

[0164] Where A and B represent the direction vectors of the two boundary lines, and θ is the fold angle. It is important to ensure that the image resolution matches the actual cable dimensions and correct for errors introduced by the shooting angle to improve calculation accuracy. The calculated fold angle serves as input data for subsequent microbend loss analysis.

[0165] Step S33: locating abnormal points of microbending loss based on the folding angle;

[0166] In this embodiment, once the fold angle is calculated, it is used as a key input for further analysis using microbend loss theory. Based on the fold angle, the cable bend radius, the refractive index of the optical fiber, and other parameters, the microbend loss is calculated using the following formula:

[0167]

[0168] Where Δα is the microbend loss, n1 is the refractive index of the optical fiber (e.g., 1.45), Δθ is the folding angle, λ is the wavelength of light (typically 1550 nm), and R is the bending radius of the optical cable. The larger the folding angle and the smaller the bending radius, the greater the microbend loss. When the calculated microbend loss exceeds a preset threshold (e.g., 0.5 dB / km), the area is considered to have a microbend loss anomaly. Through calculation, the specific location of the loss in the folding area is identified, providing data support for subsequent breakpoint location.

[0169] Step S34: Calculating the loss according to the microbending loss abnormal point to obtain microbending loss data;

[0170] In this embodiment, the accuracy of microbend loss calculation depends on the actual bending state of the optical cable. Therefore, the bend radius of each outlier must first be determined. This can be achieved through image processing techniques. Specifically, the cable image is analyzed to extract the specific shape of the folded area. During image processing, edge detection methods are used to first obtain edge information of the optical cable. This data is then used to fit the cable's curved section and determine the bend radius of each outlier. By converting the image pixel information into actual physical units, an accurate bend radius value is obtained. Using this bend radius data, combined with the cable's optical properties and fold angle, the microbend loss at each outlier is calculated. Loss calculation depends not only on the bend radius but also on the geometric characteristics of the bend area. Therefore, accurate geometric information at these points, such as the specific angle and curvature of the cable, is required. To ensure accurate calculation results, precise positioning of the bend area is crucial. This can be achieved by mapping pixel coordinates in the image to the physical world. After calculating the loss at each outlier, all loss data is aggregated to construct a loss distribution map for the cable's folded section. By combining image processing with data analysis, the loss distribution map visually displays the extent of loss at each abnormal point in the optical cable and helps analyze the overall loss of the cable. Once the loss distribution map is completed, it is compared against a preset loss threshold, typically set at 1 dB / km. All abnormal points with loss values exceeding this threshold are identified as abnormal loss areas, indicating significant damage or deterioration and requiring further investigation and repair. This microbend loss data provides important evidence for diagnosing optical cable faults. Further identification of abnormal loss points can accurately determine whether the optical cable requires maintenance or replacement. If there are numerous areas with excessive loss or if the abnormal loss is severe, the optical cable should be considered for repair or replacement. This step helps ensure the long-term stability and performance of the optical cable.

[0171] Step S35: locating the break point of the folded section optical cable data based on the microbending loss data;

[0172] In this embodiment, the break is determined by locating areas where loss data exceeds a set threshold (for example, loss greater than 2dB / km). In combination with the fiber break mechanism, the bending degree, fold angle, and stress distribution of the cable in the abnormal loss area are analyzed to identify the specific location of the potential break. The location of the break can be optimized through algorithms, such as sliding window analysis, where microbend loss data is input into the algorithm for segmented analysis to precisely locate the break point. This data provides a basis for maintenance and repair of the cable.

[0173] Step S36: Identify the optical fiber cross-sectional structural features based on the breakpoint.

[0174] In this embodiment, a scanning electron microscope (SEM) or other imaging technology is used to analyze the structural characteristics of the optical fiber cross section at the break point. The scanned image of the break point is used to identify the fracture morphology of the optical fiber, such as whether there are obvious signs of stretching or compression, or whether the optical fiber has suffered fatigue fracture at a specific point. Combined with the physical properties of the optical fiber (such as refractive index, material hardness, etc.), a detailed analysis of the break point is performed. Through cross-sectional analysis, it is further confirmed whether the optical cable has structural damage caused by microbend loss or excessive folding. This information provides detailed structural feature support for analyzing the cause of the optical cable failure.

[0175] Preferably, step S35 is specifically as follows:

[0176] Step S351: performing fiber Bragg grating measurement on the folded section optical cable data to obtain grating measurement data;

[0177] In this embodiment, grating measurement data is obtained by performing fiber Bragg grating measurement on the folded section of the optical cable. Fiber Bragg grating measurement technology can sense stress changes in the optical fiber through reflection characteristics. First, it is necessary to install Bragg grating sensors on the folded section of the optical cable. These sensors can sense wavelength changes caused by forces on the optical cable (such as bending, stretching, etc.). During installation, ensure that the fiber Bragg grating sensors are evenly distributed and that the design wavelength of the grating matches the operating wavelength of the optical cable. After the sensor is installed, a light signal of a specific wavelength is emitted by a light source. After the light signal is reflected by the Bragg fiber sensor, the wavelength change of the reflected signal is obtained using a spectrum analyzer. By measuring the change in the reflected wavelength, the strain information of the optical fiber can be directly obtained. At this time, special attention should be paid to the measurement accuracy of the fiber Bragg grating to ensure that the change in its wavelength can accurately reflect the stress change of the folded section of the optical cable. During the measurement process, the wavelength change of the fiber Bragg grating reflection signal should be recorded to obtain detailed grating measurement data.

[0178] Step S352: Calculating strain based on the grating measurement data to obtain strain data;

[0179] In this embodiment, the strain data of the optical fiber is calculated by analyzing the grating measurement data. There is a clear linear relationship between the wavelength change of the fiber Bragg grating and the strain. In this step, the strain of the optical fiber at the measurement point is first calculated by using the relationship between the wavelength change Δλ of the reflected signal and the Bragg wavelength λB, using the known optical fiber strain sensitivity (for example, the strain sensitivity of the optical fiber is usually 1.2pm / με). The calculation formula is: strain = Δλ / (λB×sensitivity), where Δλ is the change in Bragg wavelength, λB is the operating wavelength of the fiber Bragg grating, and sensitivity is a known parameter. Through this method, the strain value of each measurement point can be obtained, and the entire area of the folded section optical cable can be analyzed point by point. The generation of strain data is based on the wavelength change of each measurement point, and needs to be calibrated in combination with the actual sensitivity value to ensure the accuracy of the strain data.

[0180] Step S353: drawing a strain curve of the optical fiber according to the strain data; identifying a region where strain increases suddenly based on the strain curve of the optical fiber; and identifying a region where strain decreases suddenly based on the strain curve of the optical fiber.

[0181] In this embodiment, an optical fiber strain curve is plotted based on strain data. The abscissa of the strain curve represents the measurement location or time of the optical fiber, and the ordinate represents the strain value. When plotting the optical fiber strain curve, it is necessary to ensure the sampling accuracy and uniformity of the strain data to ensure the smoothness and accuracy of the curve. By plotting the strain data, the strain changes at different locations or times of the optical fiber can be clearly observed, thereby identifying potential abnormal areas. Then, based on the strain curve, regions of sudden strain increase and sudden strain decrease are identified. Identifying Strain Sudden Increase Regions. A strain sudden increase region refers to a region where the optical fiber experiences a significant, sharp increase in strain at a certain location or time, typically manifested as a rapid increase in strain value within a short time or distance. Strain sudden increases are typically caused by external forces or localized deformation of the optical cable at certain locations, such as bending, compression, or stretching. A strain curve with a rapidly rising slope is identified as a strain sudden increase region. To ensure accurate identification, a strain change threshold is set, for example, a strain change exceeding 0.05% is used as the threshold for a sudden increase. When the strain value increases rapidly within a certain region, and the increase exceeds the set threshold, the location is confirmed to be a strain sudden increase region. In this case, the optical cable has localized damage or mechanical stress concentration. Strain spike regions are identified based on the amplitude and slope of the curve. This refers to a sudden change in strain over a specific time or distance that exceeds a set threshold. Next, strain drop regions are identified. A strain drop region refers to a significant, sharp drop in strain at a specific location or time. This typically manifests as a rapid drop in strain over a short period of time or distance. A strain drop typically indicates stress release or structural damage to the optical fiber, resulting from fiber breakage, connector detachment, or other physical damage. A rapid drop in strain with a negative slope is considered a strain drop region. A strain change threshold is also required for identifying strain drop regions. Typically, a drop exceeding 0.05% is used as the criterion. A rapid drop in strain within a specific region, with the drop exceeding the set threshold, is considered a strain drop region. Strain drop regions are identified based on the amplitude and slope of the curve. A significant drop in strain over a short period of time or distance, with the drop exceeding the set threshold, indicates structural damage or abnormality.

[0182] Step S354: performing a region intersection operation based on the strain sudden increase region and the strain sudden decrease region to obtain a suspected fracture point region;

[0183] In this embodiment, the threshold areas for sudden increases and decreases are first defined. These areas can be defined by calculating the span and position of each sudden increase or decrease. For example, the interval of sudden increase and decrease is set to the range of 10 mm before and after the strain change is greater than the threshold. After that, the sudden increase and decrease areas are intersected to obtain the intersection area, that is, the area where the fracture occurs. This intersection area is the preliminary positioning range of the suspected fracture point. The purpose of the intersection operation is to accurately determine the fracture area of the optical cable by combining the mutation points of the strain, which helps to narrow the fault location range.

[0184] Step S355: verifying the suspected fracture point region based on the microbending loss data to obtain the fracture point region;

[0185] In this embodiment, microbend loss data is used to verify the suspected break point area and further confirm the final break point area. Microbend loss data can reflect the loss of the optical cable at the bend. If the optical cable is broken or has serious structural defects in a certain area, the microbend loss value in this area will increase significantly. Therefore, by performing microbend loss analysis on the suspected break point area, it is checked whether there is a loss value greater than a preset threshold (for example, 1dB / km). If so, the area can be further confirmed as the break point area. The specific steps are to compare the microbend loss data of the suspected break point area with the loss value of the surrounding normal area. If the difference is significant, the specific location of the break point is further confirmed. This method can improve the accuracy of break point positioning.

[0186] Step S356: Locate the break point of the folded segment optical cable data according to the break point region.

[0187] In this embodiment, the precise location of the breakpoint can be determined through a comprehensive analysis of microbend loss data, strain data, and grating measurement data in the breakpoint region. This process requires careful inspection of the folded section of the cable. Combining the optical fiber's strain curve, microbend loss distribution, and Fiber Bragg Grating (FBG) measurement results, the precise coordinates of the breakpoint can be determined step by step. In actual operation, detailed marking of the locations before and after the breakpoint is required, and the specific damage type of the cable can be further confirmed based on the cable's structural design and physical properties.

[0188] Preferably, step S36 is specifically as follows:

[0189] Step S361: Detecting brittle fracture based on the fracture point to obtain brittle fracture data; detecting ductile fracture based on the fracture point to obtain ductile fracture data; detecting microcrack propagation fracture based on the fracture point to obtain microcrack propagation fracture data;

[0190] In this embodiment, the fracture point is used as the starting position, and different types of fracture modes are identified by further analyzing the stress and strain fields around the fracture point. For brittle fracture, it is first necessary to obtain stress distribution data around the fracture point. Usually, real-time strain data is obtained through sensors such as strain gauges and fiber Bragg gratings, and analyzed in combination with the fracture toughness of the material. Brittle fracture is usually characterized by a sudden change in strain and a sharp concentration of stress, resulting in rapid crack expansion. Therefore, by analyzing the rate of change of the strain value in the area (for example, when the change exceeds 0.1%), combined with the stress concentration, the occurrence of brittle fracture can be identified. For ductile fracture, ductile failure is usually accompanied by the presence of a plastic deformation area, so it is necessary to analyze the plastic area around the fracture point, and use historical data of stress and strain, as well as the yield strength of the material to judge the characteristics of ductile fracture. Ductile fracture is usually characterized by slow crack expansion, a large deformation area, and is often accompanied by small plastic deformation (strain value exceeds 0.05%). Therefore, by comparing the strain value change trend in this area, ductile fracture can be identified. For microcrack propagation fractures, high-resolution imaging techniques (such as scanning electron microscopy or electron backscatter diffraction) are required to obtain microscopic images near the fracture point and analyze the microcrack propagation. Microcrack propagation usually manifests as microcracks or small cracks on the fracture surface, so high-resolution image processing techniques such as image segmentation and feature extraction are used to identify the area of microcrack propagation. The data of each type of fracture (brittle fracture, ductile fracture, microcrack propagation fracture) will be saved separately as different records and further integrated and analyzed through data fusion technology to facilitate subsequent fault diagnosis and maintenance decisions.

[0191] Step S362: Identifying radial cracks based on brittle fracture data;

[0192] In this embodiment, based on the brittle fracture data, the strain field and crack propagation path around the fracture point are used to identify radial cracks using an image processing algorithm. Radial cracks are usually radial cracks that extend outward from the fracture point. The direction and morphology of crack propagation are determined by analyzing the image features around the fracture area. The position of the radial crack is further confirmed by setting a crack propagation angle threshold (for example, a crack angle greater than 30° is a radial crack), and based on the texture features of the image and the width change of the crack. Radial cracks usually have a more regular shape, and the crack width changes relatively uniformly. The technology used may include edge detection, image segmentation and other processing methods, and the direction and morphology of crack propagation are determined by extracting the edge information of the image.

[0193] Step S363: detecting fibrous plastic deformation based on ductile fracture data;

[0194] In this embodiment, ductile fracture data is used in combination with the strain and deformation distribution around the fracture point to analyze whether there is fibrous plastic deformation. Ductile fracture manifests as plastic deformation over a large range, so it is necessary to locate the fibrous plastic deformation area through a strain distribution diagram. According to the size of the strain value, a strain change threshold is set (for example, a strain value greater than 0.1% is a significant plastic deformation), and the area of fibrous plastic deformation is identified by comparing the deformation characteristics of these areas. This process is judged by analyzing the changes in the strain field. Fibrous plastic deformation usually presents a long strip or filamentous structure, and the strain value of the deformation area is large.

[0195] Step S364: detecting fracture surface roughness based on microcrack propagation fracture data;

[0196] In this embodiment, based on the microcrack extension fracture data, the roughness characteristics of the fracture are detected by performing high-resolution imaging of the fracture surface near the fracture point. Microcrack extension fractures are usually manifested as irregular fracture morphology, so it is necessary to obtain a microscopic image of the fracture by scanning electron microscopy or other high-resolution imaging technology. During the image processing process, a fracture surface roughness analysis algorithm is applied, such as a roughness evaluation method based on fractal dimension or wavelet transform, to calculate the height difference and irregularity of the fracture surface. By setting a roughness threshold (for example, the average roughness of the fracture surface is greater than 2μm, which is considered to be a relatively rough fracture), the fracture roughness characteristics of microcrack extension fractures can be effectively identified.

[0197] Step S365: Integrate radial cracks, fiber-shaped plastic deformation, and fracture roughness to obtain fiber cross-sectional structural characteristics.

[0198] In this embodiment, data on radial cracks, fibrous plastic deformation, and fracture roughness are collected. The characteristics of radial cracks are usually obtained through microscopic imaging of the fracture point or crack detection on the fiber surface, recording parameters such as the crack location, length, and width; fibrous plastic deformation is obtained through strain distribution data, analyzing the degree of deformation and distribution area of the fiber after being subjected to force; fracture roughness is analyzed by high-resolution imaging equipment such as scanning electron microscopes to analyze the fracture surface, calculate the average roughness of the fracture surface and the roughness variation in different areas. Then, the data of the three are paired according to the position and time sequence of the fiber, and a data fusion algorithm is used to integrate the crack, plastic deformation, and roughness information. During the fusion process, it is necessary to perform weighting based on the morphological characteristics of the crack, the distribution range of the deformation, and the size of the fracture roughness. For example, a higher weight is given when the crack area exceeds a certain percentage, and its weight is also increased when the roughness exceeds a set standard. These data will be screened and adjusted according to preset standards. For example, when the crack width is greater than a certain set value, the area is considered to be severely damaged, and when the roughness exceeds a certain threshold (such as a roughness value exceeding 5 microns), it is considered a damaged area. Ultimately, through these weighted and standard judgments, characteristic data of the overall cross-sectional structure of an optical fiber can be obtained. These characteristics can be used for subsequent optical fiber fault diagnosis and maintenance decisions, such as whether repair or replacement of the optical fiber is needed.

[0199] Preferably, this specification also provides an optical cable defect detection system based on an image recognition model, which is used to execute the optical cable defect detection method based on an image recognition model as described above. The optical cable defect detection system based on an image recognition model includes:

[0200] The optical cable laying morphology recognition module is used to obtain an optical cable laying image; identify the optical cable laying morphology based on the optical cable laying image, and obtain the exposed section optical cable data and the folded section optical cable data;

[0201] The structural stability enhancement module is used to evaluate the corrosion resistance of the optical cable based on the exposed section optical cable data; determine the degree of armor rust based on the optical cable corrosion resistance; detect the brittleness of the optical fiber structure based on the degree of armor rust and obtain optical fiber structural brittleness data; enhance the structural stability of the exposed section optical cable data based on the optical fiber structural brittleness data and obtain structural stability data;

[0202] The optical fiber cross-sectional structure recognition module is used to construct an optical cable defect image recognition model based on optical fiber structural embrittlement data; detect microbend loss in the folded section optical cable data according to the optical cable defect image recognition model to obtain microbend loss data; locate the breakpoint of the folded section optical cable data based on the microbend loss data; and identify the optical fiber cross-sectional structure characteristics based on the breakpoint;

[0203] The optical cable structure defect detection module is used to transmit the optical fiber cross-sectional structural characteristics and structural stability data to the optical cable defect image recognition model, detect the optical cable structure defects in the optical cable laying image, and generate optical cable structure defect data.

[0204] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0205] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting optical cable defects based on an image recognition model, characterized in that: The following steps are involved: Step S1: Acquire an optical cable laying image; Identify the cable laying form based on the cable laying image to obtain the exposed section cable data and the folded section cable data; Step S2: Evaluating the corrosion resistance of the optical cable based on the exposed section optical cable data; determining the degree of armor rust based on the optical cable corrosion resistance; detecting the embrittlement of the optical fiber structure based on the degree of armor rust to obtain optical fiber structure embrittlement data; enhancing the structural stability of the exposed section optical cable data based on the optical fiber structure embrittlement data to obtain structural stability data; Step S3: constructing an optical cable defect image recognition model based on the optical fiber structure embrittlement data; detecting microbending loss of the folded section optical cable data according to the optical cable defect image recognition model to obtain microbending loss data; Locate the break point of the folded section optical cable data based on microbending loss data; Identify the structural features of the optical fiber cross section based on the break point; Step S4: Transmitting the optical fiber cross-sectional structural features and structural stability data to the optical cable defect image recognition model, detecting the optical cable structural defects in the optical cable laying image, and generating optical cable structural defect data.

2. The optical cable defect detection method based on image recognition model according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: Acquire an optical cable laying image; Step S12: removing low-frequency noise from the optical cable laying image to obtain a noise-reduced optical cable laying image; Step S13: Detecting the edge of the optical cable based on the denoised image of the optical cable laying, wherein the upper and lower detection thresholds are set to 50 and 150, respectively, to obtain the edge of the optical cable, and extracting the candidate optical cable area of the denoised image of the optical cable laying according to the edge of the optical cable; Step S14: performing an expansion operation on the optical cable candidate region using a preset 3x3 structure element to obtain an optical cable expansion region; Step S15: performing an erosion operation on the optical cable candidate area using a preset 5x5 structure element to obtain an optical cable erosion area; Step S16: Calculate the aspect ratio of the optical cable corrosion area. If the aspect ratio is greater than 2:1, obtain the exposed section optical cable data. Calculate the curvature of the corroded area of the optical cable. If the curvature radius is less than 10 cm, obtain the data of the folded section of the optical cable.

3. The optical cable defect detection method based on image recognition model according to claim 1, characterized in that: The step S2 of evaluating the corrosion resistance of the optical cable includes: Extracting surface texture data of exposed optical cable segment data; Calculating gradients based on surface texture data; Calculate the local gradient change rate based on the gradient; Determine the optical cable corrosion area of the optical cable laying image based on the local gradient change rate; Identify the corrosion depth based on the corroded area of the optical cable; Calculate the corrosion rate based on the corrosion area of the optical cable; The corrosion resistance of the exposed section of optical cable is evaluated based on the corrosion depth and corrosion rate.

4. The optical cable defect detection method based on image recognition model according to claim 1, characterized in that: Determining the rust degree of the armor layer in step S2 includes: Corrosion simulation was performed based on the corrosion resistance of the optical cable, wherein the temperature range was set to 25°C-40°C and the humidity range was set to 30%RH-98%RH to obtain corrosion simulation data; Calculate the thickness of the optical cable armor layer based on corrosion simulation data; Calculate and obtain thickness reduction data using the preset optical cable armor layer benchmark thickness and the optical cable armor layer thickness; The electrical conductivity of the corrosion simulation data was recorded using an eddy current detector; Determine corrosion depth based on electrical conductivity; The degree of rust of the armor layer is determined based on the corrosion depth and thickness reduction data.

5. The optical cable defect detection method based on image recognition model according to claim 1, characterized in that: The step S2 of detecting the embrittlement of the optical fiber structure includes: Determine the rust area according to the rust degree of the armor layer; Determine the degree of armor failure based on the rusted area; Divide the exposed cable data into armored rusted cable segments based on the degree of armor failure; The fiber bending load is simulated for the rusted armored cable section, and the fiber fatigue life is predicted to obtain the fiber fatigue life; The optical fiber structure embrittlement is detected based on the optical fiber fatigue life to obtain the optical fiber structure embrittlement data.

6. The optical cable defect detection method based on image recognition model according to claim 1, characterized in that: The step S2 of enhancing the structural stability of the exposed section optical cable data includes: Statistics of highly brittle structure optical fiber data are collected based on optical fiber structure brittle data; Determine the brittle cable section based on the exposed section optical cable data based on the highly brittle structure optical fiber data; Designing UV-resistant coating based on brittle cable segments; Laying corrosion-resistant connectors based on the brittle optical cable segment; enhancing the sealing performance of the corrosion-resistant connectors; performing laser surface treatment on the brittle optical cable segment based on the sealing performance to obtain laser surface data; performing micro-nanostructure design on the brittle optical cable segment based on the laser surface data to obtain micro-nanostructure data; The structural stability of the exposed section optical cable data is evaluated based on the micro-nano structure data to obtain the structural stability data.

7. The optical cable defect detection method based on image recognition model according to claim 1, characterized in that: Step S3 is specifically as follows: Step S31: constructing an optical cable defect image recognition model based on optical fiber structure embrittlement data; Step S32: calculating the folding angle of the folded section optical cable data according to the optical cable defect image recognition model; Step S33: locating abnormal points of microbending loss based on the folding angle; Step S34: Calculating the loss according to the microbending loss abnormal point to obtain microbending loss data; Step S35: locating the break point of the folded section optical cable data based on the microbending loss data; Step S36: Identify the optical fiber cross-sectional structural features based on the breakpoint.

8. The optical cable defect detection method based on image recognition model according to claim 7, characterized in that: Step S35 is specifically as follows: Step S351: performing fiber Bragg grating measurement on the folded section optical cable data to obtain grating measurement data; Step S352: Calculating strain based on the grating measurement data to obtain strain data; Step S353: drawing a strain curve of the optical fiber according to the strain data; Identify the strain sudden increase area based on the optical fiber strain curve; Identify the strain drop area based on the optical fiber strain curve; Step S354: performing a region intersection operation based on the strain sudden increase region and the strain sudden decrease region to obtain a suspected fracture point region; Step S355: verifying the suspected fracture point region based on the microbending loss data to obtain the fracture point region; Step S356: Locate the break point of the folded segment optical cable data according to the break point region.

9. The optical cable defect detection method based on image recognition model according to claim 7, characterized in that: Step S36 is specifically as follows: Step S361: Detecting brittle fracture based on the fracture point to obtain brittle fracture data; Detect ductile fracture based on fracture points and obtain ductile fracture data; Detecting micro-crack propagation fracture based on the fracture point to obtain micro-crack propagation fracture data; Step S362: Identifying radial cracks based on brittle fracture data; Step S363: detecting fibrous plastic deformation based on ductile fracture data; Step S364: detecting fracture surface roughness based on microcrack propagation fracture data; Step S365: Integrate radial cracks, fiber-shaped plastic deformation, and fracture roughness to obtain fiber cross-sectional structural characteristics.

10. An optical cable defect detection system based on an image recognition model, characterized in that: For executing the optical cable defect detection method based on the image recognition model according to claim 1, the optical cable defect detection system based on the image recognition model comprises: The optical cable laying morphology recognition module is used to obtain an optical cable laying image; identify the optical cable laying morphology based on the optical cable laying image, and obtain the exposed section optical cable data and the folded section optical cable data; The structural stability enhancement module is used to evaluate the corrosion resistance of the optical cable based on the exposed section optical cable data; determine the degree of armor rust based on the optical cable corrosion resistance; detect the brittleness of the optical fiber structure based on the degree of armor rust and obtain optical fiber structural brittleness data; enhance the structural stability of the exposed section optical cable data based on the optical fiber structural brittleness data and obtain structural stability data; The optical fiber cross-sectional structure recognition module is used to construct an optical cable defect image recognition model based on optical fiber structural embrittlement data; detect microbend loss in the folded section optical cable data according to the optical cable defect image recognition model to obtain microbend loss data; locate the breakpoint of the folded section optical cable data based on the microbend loss data; and identify the optical fiber cross-sectional structure characteristics based on the breakpoint; The optical cable structure defect detection module is used to transmit the optical fiber cross-sectional structural characteristics and structural stability data to the optical cable defect image recognition model, detect the optical cable structure defects in the optical cable laying image, and generate optical cable structure defect data.

Images