A method and system for measuring crack width

Abstract

The application relates to the technical field of crack detection, in particular to a crack width measurement method and system, which comprises the following steps: installing an information monitoring device in a crack to-be-detected area, obtaining crack detection initial data by using the information monitoring device; constructing a detection data optimization model, obtaining crack detection optimization data by using the detection data optimization model; establishing a detection data evaluation model, obtaining crack detection target data based on the detection data evaluation model and the crack detection optimization data; extracting crack detection position information according to the crack detection target data; and measuring the width of the to-be-detected crack according to a crack width analysis model and the crack detection position information. The application can quickly extract crack position information and accurately measure crack width by constructing data optimization, evaluation and measurement models based on crack image data.

Description

Technical Field

[0001] This invention relates to the field of crack detection technology, specifically to a method and system for measuring crack width. Background Technology

[0002] With the rapid development of technologies such as computer vision, image processing, pattern recognition, artificial intelligence, and digital signal processing, the application scenarios of video surveillance technology are becoming increasingly diverse. Among them, intelligent video surveillance technology is regarded as the future development direction of the industry. Existing crack detection technologies and instruments cannot yet meet the comprehensiveness and accuracy required for practical applications, especially in terms of monitoring data optimization or crack measurement and tracking analysis under different external interference conditions.

[0003] Currently, the analysis of crack conditions mainly relies on intelligent monitoring instruments and manual recording of information. However, the above methods are clearly insufficient in terms of objectivity, accuracy, and authority. In addition, traditional crack detection methods are not thorough enough in data processing, and the detection work is cumbersome, inaccurate, and inefficient.

[0004] Therefore, it is necessary to optimize existing crack width detection methods to achieve efficient searching, accurate acquisition, precise measurement, and in-depth analysis of crack width measurement technology, and further track the changing trend of cracks over a certain period of time, thereby improving the efficiency and accuracy of crack detection and providing more reliable and accurate data support for related industries. Summary of the Invention

[0005] To address the shortcomings of existing methods and meet the needs of practical applications, this invention utilizes a dynamic camera to collect data from the crack detection area and employs data optimization and evaluation models to enhance the accuracy of the detection data. This allows for better measurement of target cracks in complex and dynamic environments, ultimately achieving accurate measurement and objective analysis of crack width. Specifically, this invention provides a crack width measurement method, comprising the following steps: installing an information monitoring device in the crack detection area to obtain initial crack detection data; constructing a detection data optimization model to obtain optimized crack detection data; establishing a detection data evaluation model to obtain target crack detection data based on the evaluation model and the optimized crack detection data; extracting crack detection location information based on the target crack detection data; and measuring the width of the crack based on the crack width analysis model and the crack detection location information. This invention, based on the target crack detection data, accurately extracts the crack location information, which helps in accurately analyzing crack conditions and is of great significance for assessing crack severity, developing repair plans, and predicting crack development trends.

[0006] Optionally, installing information monitoring equipment in the crack detection area and obtaining initial crack detection data using the information monitoring equipment includes: installing information monitoring equipment in the crack detection area, wherein the information monitoring equipment includes multiple dynamic cameras. The cameras of this invention can capture crack changes and related data in real time, which helps to understand crack changes over different time periods and enables a more accurate assessment of the nature and severity of the crack.

[0007] Optionally, the step of constructing a detection data optimization model and obtaining optimized crack detection data through the detection data optimization model includes: obtaining the pixel ratio of the image to be detected based on the initial crack detection data; establishing a detection data optimization model based on the pixel ratio; and obtaining optimized crack detection data using the detection data optimization model. This invention constructs a detection data optimization model and obtains pre-optimized crack detection data through this model, which can significantly improve data quality, enhance crack feature extraction, reduce computational complexity, improve model generalization ability, and provide a solid foundation for subsequent analysis.

[0008] Optionally, the detection data optimization model satisfies the following relationship:

[0009]

[0010] in, This represents the optimized image. This represents the minimum output value of the converted image. This represents the grayscale pixel ratio of the image to be detected. This represents the maximum output value of the converted image. The detection data optimization model of this invention can be adjusted and optimized according to different situations, thereby better extracting crack features, eliminating interference factors, controlling the output range, improving automation and computational efficiency, and helping to improve the accuracy and efficiency of crack detection methods.

[0011] Optionally, establishing a detection data evaluation model and obtaining crack detection target data based on the detection data evaluation model and the crack detection optimization data includes: constructing a detection data evaluation model using the crack detection optimization data; and obtaining data evaluation results of the image to be detected through the detection data evaluation model. The crack detection target data obtained by this invention based on the data evaluation model can make the data closer to the real situation, reducing the possibility of false detections and missed detections, thus ensuring the relevance and effectiveness of the measurement work.

[0012] Optionally, the detection data evaluation model satisfies the following relationship:

[0013]

[0014] in, This represents the evaluation result of the dataset of images to be detected. This indicates the total number of images to be detected. This represents the initial data set of the image to be detected. This represents the dataset used as a reference for the detected image. This represents the fluctuation coefficient of the data acquisition equipment. This indicates the weight of external influences on the monitored object.

[0015] This invention comprehensively considers multiple factors to fully assess data quality, accurately reflect the actual situation, optimize data processing flow, provide decision support, and enhance the model's versatility and adaptability.

[0016] Optionally, obtaining the crack detection target data based on the detection data evaluation model and the crack detection optimization data includes: evaluating and analyzing the crack detection optimization data using the detection data evaluation model and obtaining data evaluation results; and combining the data evaluation results and the detection data optimization model to obtain the crack detection target data. This invention comprehensively applies evaluation and optimization models to more accurately reflect the true condition and attribute characteristics of cracks, helping to reduce the possibility of false detections and missed detections, and improving the accuracy of crack detection results.

[0017] Optionally, extracting crack detection location information based on the crack detection target data includes: constructing a crack location correction model; and obtaining crack detection location information based on the crack location correction model and the crack detection target data. This invention constructs a crack location correction model and extracts crack detection location information based on this model and the crack detection target data, which can improve the accuracy of location information, optimize the data processing flow, enhance the robustness of the system, and provide reliable support for measurement work.

[0018] Optionally, measuring the width of the crack to be tested based on the crack width analysis model and the crack detection location information includes: constructing a crack width analysis model; the crack width analysis model satisfies the following relationship;

[0019]

[0020] in, This represents the distance between any two detection points in the image to be detected. This represents the first dimension in the horizontal direction of the first detection point in the image to be detected. This represents the first dimension in the horizontal direction of the second detection point in the image to be detected. This represents the second dimension in the horizontal direction of the first detection point in the image to be detected. This represents the second dimension in the horizontal direction of the second detection point in the image to be detected. This indicates that the first detection point in the image to be detected is perpendicular to the third dimension of the image's horizontal direction. This represents the third dimension of the second detection point in the image, perpendicular to the horizontal direction of the image. This invention calculates the distance between any two points in the target image based on a model, more accurately determining the width of the crack, thus providing a more comprehensive reflection of the actual width of the crack.

[0021] Secondly, to efficiently execute the crack width measurement method provided by this invention, this invention also provides a crack width measurement system, including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to call the program instructions to execute the crack width measurement method as described in the first aspect of this invention. The crack width measurement system of this invention has a compact structure and stable performance, and can stably execute the crack width measurement method provided by this invention, improving the overall applicability and practical application capability of this invention. Attached Figure Description

[0022] Figure 1 This is a flowchart of the crack width measurement method of the present invention;

[0023] Figure 2 This is a schematic diagram of the abnormal variation of the crack width measurement method detection equipment of the present invention;

[0024] Figure 3 This is a flowchart of the crack location information extraction method for crack width measurement according to the present invention;

[0025] Figure 4 This is a schematic diagram comparing the data errors of different measurement methods for the crack width measurement method of the present invention;

[0026] Figure 5 This is a structural diagram of the crack width measurement system of the present invention. Detailed Implementation

[0027] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0028] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0029] Please see Figure 1 This invention utilizes a camera to acquire real-time image data of the crack area to be measured. By employing data optimization and evaluation models, the acquired data is finely processed and evaluated, significantly improving the accuracy of the detection data. This allows for efficient and accurate measurement of the target crack, ensuring the reliability of the crack width measurement results. The invention provides a crack width measurement method, which includes the following steps:

[0030] S1. Set up information monitoring equipment in the crack detection area and use the information monitoring equipment to obtain the initial data of crack detection. The specific implementation steps and contents are as follows.

[0031] Information monitoring equipment is set up according to the crack detection area. In this embodiment, the information monitoring equipment includes multiple dynamic cameras, and its specific implementation is as follows:

[0032] First, the area to be tested for cracks needs to be analyzed, including but not limited to the crack's shape, area, terrain, and activity patterns. Based on these characteristics and actual needs, a suitable dynamic camera model should be selected. Since different cameras differ in performance and functionality, key parameters such as the number of cameras, deployment locations, and frame rate need to be set according to the actual conditions of the area to be tested.

[0033] When selecting a camera model, the various parameters of the camera are finely configured and adjusted according to the actual situation of the area to be tested and the data detection requirements. These parameters include, but are not limited to, frame rate, resolution, color performance, brightness, contrast and storage space, to ensure that the acquired crack detection images have excellent quality and clarity.

[0034] In addition, factors such as the angle, range, and height of crack detection will be comprehensively considered, and the dynamic positions of multiple cameras will be carefully selected and designed to ensure that the coverage and clarity of the detection equipment are at their best. At the same time, it will be ensured that each camera can accurately capture the key parts of the crack. Special attention will be paid to the monitoring height and angle of the cameras when deploying them to avoid data acquisition blind spots as much as possible, so as to ensure that the best detection images of the cracks to be measured can be obtained. This will further realize comprehensive, efficient, and accurate monitoring of the crack detection area, providing a solid foundation for subsequent data analysis and processing.

[0035] Next, multiple cameras will be connected via signal cables to ensure stable communication between the cameras and the data storage device. Each camera will then undergo further debugging with the data storage system to ensure accurate transmission of detection data. During debugging, key parameters such as camera focus, aperture, and color will be adjusted to optimize image clarity and improve data accuracy. Simultaneously, to ensure the integrity and traceability of monitoring data, the data storage device is designed with a data storage scheme to securely store all detection data, facilitating subsequent data preservation and retrieval.

[0036] Furthermore, to ensure the normal operation of the data storage device, the detection data is analyzed and processed regularly to promptly identify and resolve any potential problems. This ensures the proper functioning of the entire crack detection method and provides reliable data support for crack width monitoring and analysis.

[0037] Furthermore, the camera monitoring method and data acquisition method in this embodiment are only optional conditions of the present invention. In other embodiments, the detection equipment can be flexibly selected according to actual needs. Through the collaborative work of multiple cameras, the effective detection of the target to be measured can be achieved, the crack width and changes can be analyzed more accurately, the accuracy of the detection data can be improved, and the practicality and accuracy of the crack width measurement method can be enhanced.

[0038] S2. Construct a detection data optimization model to obtain optimized crack detection data through the detection data optimization model. The specific implementation steps and contents are as follows:

[0039] Based on the initial data analysis of crack detection, the pixel ratio of the crack image to be detected is analyzed to explore the proportion of specific gray-level pixels in the detected image, so as to more comprehensively understand the relevant characteristics of the crack image. In the embodiment, the pixel ratio analysis satisfies the following relationship:

[0040]

[0041] in, This represents the proportion of a specific grayscale pixel in the image to be detected. This indicates the stability coefficient of the testing equipment. This represents the number of image pixels that satisfy a specific range of grayscale values. This represents the total number of pixels in the image to be detected.

[0042] The specific grayscale pixel ratio value refers to the proportion or ratio of a specific range of grayscale values ​​in a crack detection image. A grayscale image is an image that contains only brightness information and no color information. Its pixel value range is usually from black to white, representing different brightness levels. In this embodiment, the grayscale pixel ratio value is obtained based on the ratio of the number of specific pixels in the image to the total number of pixels in the image. It reflects the proportion of specific grayscale pixels in the image and further provides information such as the brightness distribution, contrast, and noise of the crack detection image.

[0043] The stability coefficient of a testing device is an important indicator of its ability to maintain its posture or operational stability under specific environmental conditions. It reflects the degree to which the system maintains stability and balance when subjected to external forces or environmental changes. A higher stability coefficient indicates better stability and the ability to guarantee accurate and reliable measurement results under various environmental conditions. Therefore, the stability coefficient is a key factor to consider when selecting and using testing equipment to ensure the accuracy and reliability of the test results.

[0044] In crack detection applications, the number of image pixels meeting a specific range of grayscale values ​​refers to the number of pixels in a grayscale image whose grayscale values ​​fall within that specific range. This specific range is usually determined based on empirical values ​​of crack grayscale, aiming to effectively distinguish cracks from background areas. Firstly, when the difference between crack grayscale and background is significant, a global thresholding method can be used. In an optional embodiment, a maximum grayscale value (max_crack_gray) is set, representing the maximum grayscale value expected the crack can reach. In this method, all pixels in the image with grayscale values ​​less than max_crack_gray are considered crack areas, while pixels with grayscale values ​​greater than or equal to max_crack_gray are considered background areas. By counting the number of pixels with grayscale values ​​less than max_crack_gray, the number of pixels in the crack area can be obtained.

[0045] When the gray level of the crack is similar to that of the background, a global thresholding method is used. In this case, it is necessary to calculate the average gray level or Gaussian weighted gray level of the local neighborhood of each pixel in the image and compare it with a dynamic threshold, so as to more accurately identify the crack region based on the characteristics of the pixel neighborhood.

[0046] Furthermore, if more precise control over the distribution of grayscale values ​​is required, a grayscale mapping method can be used. In another alternative embodiment, the grayscale values ​​of the original image are mapped to a new grayscale value range to ensure that the grayscale values ​​of the cracks fall within the specified range. By adjusting the mapping function, the contrast between the cracks and the background can be enhanced, making it easier to identify the crack area.

[0047] Based on this, the number of pixels in the crack detection grayscale image that meet a specific range of grayscale values ​​can be obtained. This number of pixels provides important information about the size and distribution of the crack region, which is helpful for subsequent crack analysis and processing.

[0048] Each pixel in a grayscale image has a grayscale value, which represents the brightness of the corresponding pixel. Monitoring a certain brightness range or feature of an image can be calculated by the number of pixels within a specific grayscale value range, which helps in subsequent image analysis, processing, and recognition processes.

[0049] The total number of pixels in the image to be detected refers to the total number of pixels in the crack detection image. In digital image processing, an image is composed of pixels, and each pixel is the smallest unit of an image, containing information including but not limited to color and brightness. Therefore, the total number of pixels in the image to be detected is the sum of all the pixels that make up the image. Furthermore, the number of pixels is related to the image resolution; the higher the resolution, the more pixels there are, and the richer the image details. Based on the total number of pixels in the image to be detected, we can understand the size and complexity of the image, providing fundamental data for subsequent crack image processing and analysis.

[0050] Then, based on the pixel ratio of the image to be detected, a detection data optimization model is established to obtain accurate information about the crack image to be detected. The specific implementation details are as follows:

[0051] First, a non-linear transformation of the grayscale to be detected is required, which means non-linearly mapping or adjusting the grayscale levels in the crack detection image. The main purpose of image transformation is to expand or compress the grayscale range of the image in order to better display the details of the crack detection image or adapt to specific display requirements.

[0052] In an optional embodiment, when the grayscale display range of the display is limited, directly displaying some images with a large grayscale value range may result in the loss of some details or insufficient contrast. Therefore, performing logarithmic transformation and nonlinear operation on the grayscale values ​​can map the originally wide grayscale range into a narrower range, thereby enabling the display to perform better.

[0053] In this embodiment, the logarithmic transformation described above needs to satisfy the following relationship:

[0054]

[0055] in, This represents the output result of the transformed image to be detected. This represents the grayscale scaling factor of the image to be detected. This represents the proportion of a specific grayscale pixel in the image to be detected.

[0056] The converted image output refers to the image data or representation obtained after performing a non-linear grayscale transformation on the image to be detected. During the conversion process, the grayscale values ​​of the original image are adjusted and processed, and the image grayscale is converted through logarithmic transformation, power law transformation, or other non-linear transformation methods.

[0057] The grayscale scaling factor of the image to be detected refers to a scaling factor used to adjust or change the range of grayscale values ​​in the image processing. This factor determines the degree to which grayscale values ​​increase or decrease, thereby adjusting the brightness and contrast of the image to be detected. Different scaling factors can stretch or compress the grayscale range of the image, optimizing the visual effect or adapting it to specific processing needs.

[0058] In this embodiment, the value of the scaling factor can be determined based on the original grayscale distribution of the image to be detected, the processing purpose, and the expected effect. A larger scaling factor usually leads to a larger change in grayscale values, thereby enhancing the contrast of the image; while a smaller scaling factor may produce a more gentle adjustment to the grayscale values, which can maintain the softness and detail of the image.

[0059] In this embodiment, based on the logarithmic transformation formula and combined with the proportion of different grayscale pixels in the image to be detected, the maximum and minimum output values ​​of the image to be detected are analyzed in depth. This helps to more accurately understand the grayscale distribution characteristics of the image and provides strong support for subsequent processing and analysis.

[0060] The maximum output value of the image after grayscale nonlinear transformation refers to the maximum brightness value that the image to be detected can achieve in grayscale after logarithmic transformation. In this embodiment, the maximum output value reflects the performance of the brightest pixel in the image after logarithmic transformation, and is closely related to the image's contrast, brightness distribution, and the settings of the transformation parameters. Based on the maximum output value, we can understand the brightness range and distribution of the image after transformation, thereby further evaluating the impact of logarithmic transformation on image quality and whether it meets specific application requirements. In crack detection applications, changes in the maximum output value reveal the brightness characteristics of the crack region, helping to more accurately identify and analyze the morphology and distribution of cracks.

[0061] The minimum output value of an image after a nonlinear grayscale transformation refers to the minimum brightness value that the image to be detected can achieve in grayscale after logarithmic transformation. The minimum output value reflects the performance of the lowest-brightness pixel in the image after the logarithmic transformation, and it is related to the image's dark detail, noise level, and the settings of the transformation parameters. In practical applications of crack detection, changes in the minimum output value reveal detailed information about dark areas in the image, including but not limited to the crack's starting point, depth, or width. By analyzing the minimum output value to compare the preservation of dark details in the transformed image, crack features can be identified and extracted more accurately. It also helps in evaluating the contrast of the detected image and highlighting crack edges.

[0062] Based on the analysis results of the grayscale pixel ratio and the nonlinear transformation results of the image to be detected, the image to be detected is optimized. In this embodiment, the grayscale value distribution of pixels in the image to be detected is mainly adjusted to make it approach a uniform distribution. In an optional embodiment, the originally unevenly distributed image histogram is made approximately uniform, thereby improving the image contrast. After equalization optimization, the pixels in the image can make fuller use of each grayscale level, making the grayscale distribution more uniform, significantly improving the dynamic range and contrast of the image to be detected, enhancing the contrast and visual effect of the image, and making the image to be detected more vivid, clear and accurate.

[0063] In this embodiment, the above-mentioned detection data optimization model satisfies the following relationship:

[0064]

[0065] in, This represents the grayscale pixels of the optimized image to be detected. This represents the minimum output value of the image to be detected after grayscale nonlinear transformation. This represents the grayscale pixel ratio of the image to be detected. This represents the maximum output value of the image to be detected after grayscale nonlinear transformation.

[0066] By using the detection data to optimize the model and redistribute the pixel values ​​of the image to be detected, the details in both the dark and bright parts of the image can be better displayed, thereby significantly enhancing the overall contrast of the image, making the details of the image to be detected more obvious, and improving the data quality. On the other hand, it helps to improve the image quality and readability, thereby improving the accuracy and robustness of the data optimization algorithm.

[0067] Next, the crack detection optimization data is obtained by optimizing the model using the detection data.

[0068] The initial crack detection data is optimized using a data optimization model to obtain more accurate optimized crack detection data. In one optional embodiment, the initial crack detection data is first preprocessed, including but not limited to data cleaning, noise reduction, and standardization, to ensure the accuracy and consistency of the initial data. Then, the initial data is input into the crack detection model to obtain optimized crack detection data. Furthermore, based on the initial data and optimization results, the data optimization model is iteratively trained, its parameters are adjusted, and the algorithm is optimized to better identify and process the initial crack data. Finally, through this series of optimization operations, pre-optimized crack detection data is obtained. This data not only has higher accuracy and reliability but also better reflects the actual situation of the cracks, providing strong support for subsequent crack analysis and processing.

[0069] Furthermore, the initial data optimization processing method in this embodiment is merely an optional condition of the present invention. In other embodiments, the detection data optimization method can be changed according to the data collection situation and target requirements to improve the accuracy, reliability and readability of the detection data, while improving the actual performance of the data processing model.

[0070] S3. Establish a detection data evaluation model. Based on the detection data evaluation model and the pre-optimized crack detection data, obtain the target crack detection data. The specific implementation steps and contents are as follows:

[0071] A detection data evaluation model is established based on crack detection optimization data, and the data evaluation results of the image to be detected are analyzed through the detection data evaluation model. The specific implementation content is as follows:

[0072] By utilizing crack detection optimization data, a detection data evaluation model was further constructed. This evaluation model can comprehensively assess and analyze the optimized data of the images to be detected, thereby ensuring the accuracy and validity of crack-related data. By applying the detection data evaluation model, the crack situation in the images to be detected can be more accurately determined, providing a scientific basis for subsequent crack monitoring and measurement, thus guaranteeing the accuracy and reliability of crack detection results.

[0073] In this embodiment, the detection data evaluation model satisfies the following relationship:

[0074]

[0075] in, This represents the evaluation result of the dataset of images to be detected. This indicates the total number of images to be detected. This represents the initial data set of the image to be detected. This represents the dataset used as a reference for the detected image. This represents the fluctuation coefficient of the data acquisition equipment. This indicates the weight of external influences on the monitored object.

[0076] The evaluation result of the image dataset to be detected refers to the comprehensive assessment conclusion obtained after a full analysis of the dataset using a detection data evaluation model. This result primarily covers the crack detection performance of the images in the dataset, including but not limited to the accuracy, completeness, and reliability of information such as crack location, shape, and size. Furthermore, the evaluation result comprehensively considers indicators such as crack recognition rate, false detection rate, and false negative rate for all images in the dataset, as well as factors such as image clarity, contrast, and noise level. A comprehensive analysis of these indicators and factors is conducted to determine the overall performance of the image dataset in crack detection.

[0077] The total number of images to be detected refers to the total number of images to be processed, analyzed, or evaluated in this embodiment. These images are typically collected based on specific application requirements and actual conditions. In practical applications of crack detection, the number of images involves multiple aspects, including but not limited to images acquired within different time periods, images from different locations or regions, and images of different resolutions or formats.

[0078] The initial image dataset refers to the raw image data collected in step S1 before crack detection or other related image processing tasks. This includes, but is not limited to, crack images taken from different angles, distances, and lighting conditions, as well as images of cracks with various background and interference factors. These images have different resolutions, color spaces, and formats, thus reflecting the specific appearance and informational characteristics of the crack to be measured under different conditions.

[0079] The reference dataset for the detection images refers to the optimal set of image data used to achieve crack detection. In this embodiment, the dataset is obtained by further optimization and organization based on the initial image dataset, which can meet the specific needs of the crack measurement task. In this embodiment, the relevant requirements and conditions of the reference image dataset can be set in advance according to the crack detection area, the initial data status, and the width measurement target, so that the algorithm or model can better optimize the crack data, thereby achieving accurate and efficient crack width detection.

[0080] The volatility coefficient of a data acquisition device is a numerical indicator representing the degree of data variation during the data acquisition process. It can be used to assess the relative dispersion of the data. The larger the volatility coefficient, the greater the degree of data variation, meaning that the data acquisition device may experience significant fluctuations or instability during the acquisition process. Conversely, the smaller the volatility coefficient, the smaller the relative dispersion of the data, the more stable the data distribution, and the more stable the performance of the data acquisition device.

[0081] The volatility coefficient can help in better understanding and evaluating the performance of data acquisition equipment, as well as the accuracy and reliability of data acquisition results. A high volatility coefficient may necessitate checking the equipment's operating status, environmental conditions, or data acquisition methods, and making corresponding optimizations or adjustments. Furthermore, the volatility coefficient can serve as a risk assessment parameter, helping to predict and avoid potential problems, thus enabling the development of a more reasonable and reliable initial data acquisition plan.

[0082] The weighting of external influences on a monitored object specifically refers to the degree or proportion of influence of external factors on the state or performance of the monitored object. During crack measurement, the monitored object, including but not limited to buildings and bridges, is affected by various external factors, including but not limited to wind, temperature, humidity, vibration, and load. These factors can all cause deformation, stress changes, or material property degradation in the cracked structure, thus affecting the generation, development, and distribution of cracks. Quantifying the weighting of external influences can reflect the impact of different external factors on cracks, thereby enabling the development of better data acquisition schemes and width measurement methods.

[0083] Then, the initial crack detection data is evaluated and analyzed using the detection data evaluation model, and the data evaluation results are obtained.

[0084] Each image in the crack detection data was analyzed individually using an evaluation model to obtain evaluation results for the entire dataset. Crack features were extracted from the images and compared with existing standard features. A comprehensive evaluation of the crack detection data images was then conducted, incorporating evaluation results, data feature matching degree, false positive rate, and false negative rate. This evaluation model ensured the accuracy and effectiveness of the crack detection data, laying a solid foundation for subsequent crack detection work.

[0085] Finally, the target data for crack detection is obtained by combining the data evaluation results and the detection data to optimize the model.

[0086] Comparing and analyzing the data evaluation results with existing high-quality data and standard indicator data, and identifying the deficiencies and defects in the initial crack detection data through comparison, not only helps to deeply understand the practical performance of crack detection data, but also provides clear guidance for subsequent data optimization.

[0087] Comparative analysis reveals discrepancies in the initial data, including but not limited to key indicators such as the accuracy of data feature extraction, false positive rate, and false negative rate. Based on the comparison results and the detection data optimization model, the crack detection data is iteratively optimized. This iterative optimization process includes, but is not limited to, adjusting the parameters of the data optimization model, improving feature extraction methods, and optimizing the data processing flow, thereby enhancing the accuracy and reliability of the crack detection data.

[0088] In an alternative embodiment, the differences between the dataset evaluation results and the standard reference dataset are analyzed to determine the direction of parameter adjustments that need to be made.

[0089] Although the total number of images is not a parameter that can be directly adjusted, it can affect the efficiency and accuracy of iterative optimization. More images can provide more data support, helping the algorithm to better learn and adapt to crack features. However, too many images may also increase the computational burden and complexity. Therefore, in practical applications, the total number of images needs to be selected reasonably based on computational resources and time costs.

[0090] The fluctuation coefficient of data acquisition equipment can be used to compensate for the impact of equipment instability on image quality. Due to factors such as equipment aging, temperature changes, or mechanical vibration, data acquisition equipment may experience fluctuations. In practical applications, adjusting the fluctuation coefficient value based on comparison results can reduce the impact of fluctuation factors on crack detection results and improve the accuracy of the algorithm.

[0091] The external influence weight of the monitored object is used to consider the influence of external environmental factors, such as light, shadow, and occlusion, on crack characteristics. Different application scenarios and environmental conditions require different external influence weights. By adjusting the external influence weights, the algorithm can better adapt to different environmental conditions and improve the accuracy and robustness of crack detection.

[0092] During the iterative optimization process, the adjustment of model parameters needs to be comprehensively considered based on specific task requirements, data characteristics, and algorithm performance. Through continuous trial and adjustment, the optimal parameter combination can be found to achieve more accurate crack detection and more efficient image processing.

[0093] In an optional embodiment, the gap between the initial data and high-quality data or standard indicators can be gradually narrowed through multiple iterative optimizations, ultimately obtaining the target data for crack detection. This target data not only has higher accuracy and meets the needs of practical applications, but also provides solid data support for subsequent crack monitoring and treatment.

[0094] In this embodiment, the data evaluation results are compared and analyzed with high-quality data and standard index data, and iterative optimization is carried out in combination with the detection data optimization model. This is to obtain high-quality crack detection target data, which not only improves the performance of crack detection data, but also provides a reference for the continuous optimization of crack detection algorithms and models.

[0095] Furthermore, the method for acquiring crack detection target data in this embodiment is merely an optional condition. In other embodiments, the data optimization method can be adjusted according to data requirements. By making corresponding adjustments to the data optimization method based on specific data requirements, the applicability and flexibility of the crack width measurement method can be improved.

[0096] S4. Extract crack detection location information based on crack detection target data. The specific implementation steps and contents are as follows:

[0097] In crack detection methods, cameras or other types of image acquisition devices can be used to establish a coordinate system centered on the camera. This coordinate system can describe the correspondence between the camera and the surface of the crack to be detected or other objects to be detected.

[0098] In this embodiment, the image acquisition device, i.e., the camera in this embodiment, has its coordinate system origin defined at the optical center of the camera, i.e., the center point of the lens. The X-axis is typically parallel to the horizontal direction of the image sensor, with its positive direction pointing to the right side of the image sensor; the Y-axis is parallel to the vertical direction of the image sensor, with its positive direction pointing downwards from the image sensor; and the Z-axis is perpendicular to the plane of the image sensor, pointing in front of the camera lens, i.e., the shooting direction. Based on this, a correspondence can be established with the object being detected. Through the camera coordinate system, the specific location of each pixel in the image on the crack surface or other object being detected in the real world can be clearly determined.

[0099] In this embodiment, establishing a coordinate system centered on the camera is one step in the crack detection method. This helps to accurately describe the correspondence between the camera and the surface of the crack to be detected, or other objects to be detected, and provides important spatial information for subsequent crack detection and localization. However, if the manufacturing and assembly process of the camera lens is not precise enough or contains process deviations, distortion may occur, resulting in distorted images and an inability to accurately reproduce the actual scene. For information on distortion variations in the detection equipment, please refer to [link to relevant documentation]. Figure 2 Where A represents the actual location of the crack to be detected, A1 represents the ideal imaging location of the crack to be detected, A2 represents the radial distortion location of the crack to be detected, and A3 represents the tangential distortion location of the crack to be detected.

[0100] Tangential distortion in the detection equipment mainly originates from the camera assembly process, which results in the failure to ensure that the lens and imaging surface are strictly parallel during assembly. Radial distortion is a typical lens distortion phenomenon, characterized by distortion distributed along the radius of the lens. This distortion is caused by the fact that the refraction of light in the central part of the lens is more curved than that in the edge part.

[0101] Furthermore, radial distortion can be divided into barrel distortion and pincushion distortion. Barrel distortion is characterized by the shape of the central region of the image bulging outward, like an inverted barrel; while pincushion distortion is the opposite, with the shape of the central region of the image concave inward, resembling a pillow. Both of these distortion forms are clearly manifested in crack detection images.

[0102] In an optional embodiment, when the distortion at the center of the optical axis reaches a minimum value, i.e., close to 0, the distortion becomes increasingly pronounced as the lens radius gradually increases, specifically manifested as an increase in image distortion. To reduce the impact of device distortion on crack detection data, relevant adjustment parameters are introduced to optimize the crack's location and restore the actual condition of the crack to be detected.

[0103] A crack location correction model is constructed based on crack detection data. The crack image coordinates are adjusted using this model to obtain a crack image that more closely resembles the real scene. The image coordinate correction model satisfies the following relationship:

[0104]

[0105] in, This represents the first horizontal dimension of the image after calibration. This represents the first dimension of the image in the horizontal direction. This indicates the first adjustment weight. This indicates the lens radius value of the detection equipment. This indicates the second adjustment weight. This indicates the third adjustment weight. This represents the second horizontal dimension of the image after correction. Represents the second dimension of the image in the horizontal direction. This represents the third dimension perpendicular to the horizontal direction of the image after calibration. This represents the third dimension, which is perpendicular to the horizontal direction of the image.

[0106] The image dimension after calibration refers to the new position coordinates of each pixel in the image after distortion correction. During the calibration process, the coordinates of the crack image are adjusted according to the distortion mathematical model to eliminate image distortion caused by lens distortion. Based on this, pixels that were originally offset due to distortion are repositioned to their proper positions, resulting in a more accurate and distortion-free image. This accurately reflects the spatial relationship of the crack image, allowing objects and scenes in the crack detection image to be presented more realistically, improving the accuracy and reliability of the image processing results.

[0107] Image coordinate adjustment weights refer to assigning different adjustment coefficients or weight values ​​to different pixels or regions during image coordinate calibration, based on their degree of distortion. Because camera distortion varies at different locations or in different directions, different weights need to be set according to the degree of distortion of each pixel or region during image coordinate adjustment to achieve more accurate position correction.

[0108] In an optional embodiment, the image adjustment weights need to be determined based on factors such as the distortion model of the detection device, lens characteristics, and actual shooting conditions. For areas with more severe distortion, a larger adjustment weight value is assigned to allow for more significant coordinate adjustments; conversely, for areas with less distortion, a smaller adjustment weight value is used to avoid overcorrection. By reasonably setting the image adjustment weights, differentiated processing of distortion in different areas of the image can be achieved, thereby improving the correction effect of the entire detection image. This helps to restore the true shape and size of the crack detection image, improves the quality of the detection image, and provides a more accurate data basis for subsequent crack measurements.

[0109] Based on the crack location correction model and crack detection target data, crack detection location information is extracted. The specific implementation details are as follows:

[0110] Crack location information is extracted using crack detection target data. During this extraction process, features such as crack morphology, length, and width must be considered, as these features are crucial for analyzing crack stability and safety. Therefore, appropriate algorithms and techniques are needed to accurately measure and describe crack-related features. The extracted crack location information is then combined with a calibration model to further refine and optimize the crack location, resulting in more accurate and reliable crack location information. This provides support for subsequent crack width measurement. For a detailed explanation of the crack detection location information extraction process, please refer to [link to relevant documentation]. Figure 3Where (x1, y1) represents the coordinates of the first pixel of the crack image, (x11, y11) represents the coordinates of the first pixel of the crack image after dimensional processing, (x1, y1, z1) represents the three-dimensional coordinates of the first pixel of the crack image, (x2, y2) represents the coordinates of the second pixel of the crack image, (x21, y21) represents the coordinates of the second pixel of the crack image after dimensional processing, (x2, y2, z2) represents the three-dimensional coordinates of the second pixel of the crack image, c1 represents the crack image feature matching program, c2 represents the crack image pixel coordinate dimensional processing program, c3 represents the crack image pixel three-dimensional coordinate calculation program, and S represents the crack detection location information set. Dimensional processing refers to optimizing the coordinates of each pixel of the crack image using an image coordinate calibration model.

[0111] Furthermore, the analysis steps and processing methods for crack location information in this embodiment are merely optional conditions. In other embodiments, the method for obtaining crack location data can be optimized according to the implementation conditions and actual data, which not only improves the adaptability and flexibility of the crack width measurement method, but also enhances data processing efficiency and accuracy.

[0112] S5. Measure the width of the crack to be tested based on the crack width analysis model and crack detection location information. The specific implementation steps and contents are as follows:

[0113] In this embodiment, based on the crack formation mechanism and influencing factors, appropriate prediction theories and simulation methods are selected to establish a crack width analysis model. Based on the above theories and methods, an analysis model for crack width is established. At the same time, a portion of crack target detection data is randomly selected to train and verify the analysis model. The prediction accuracy and reliability of the analysis model are verified based on the training results to improve the performance of the analysis model.

[0114] The above crack width analysis model satisfies the following relationship;

[0115]

[0116] in, This represents the distance between any two detection points in the image to be detected. This represents the first dimension in the horizontal direction of the first detection point in the image to be detected. This represents the first dimension in the horizontal direction of the second detection point in the image to be detected. This represents the second dimension in the horizontal direction of the first detection point in the image to be detected. This represents the second dimension in the horizontal direction of the second detection point in the image to be detected. This indicates that the first detection point in the image to be detected is perpendicular to the third dimension of the image's horizontal direction. This represents the third dimension of the second detection point in the image to be detected, which is perpendicular to the horizontal direction of the image.

[0117] In this embodiment, two or more detection points are randomly selected in the image of the crack to be detected, and then the distance between the detection points is calculated based on the detection points. The distance of the crack to be detected is analyzed based on the distance between different detection points.

[0118] The distance between any two detection points in the image to be detected refers to the spatial distance between two randomly selected crack images or feature points as detection points in the image to be detected. This helps to quickly understand the shape, size, and positional relationship of cracks in the image to be detected, and then measure the actual distance value of the cracks.

[0119] Furthermore, based on the calibrated and verified crack location data and crack width analysis model, the width of the crack to be detected can be accurately obtained. The output of the above model not only reflects the actual width of the crack, but also provides an important basis for quantitative analysis and evaluation of the crack condition.

[0120] Furthermore, to verify the beneficial effects and practical value of the crack width measurement method of the present invention, a comparative analysis is now conducted between the crack width measurement method proposed in this invention and existing measurement methods. The specific implementation details are as follows:

[0121] Real-world cracks exhibit diverse tilt angles, making it difficult for detection equipment to ensure perfect parallelism with the crack's surface during image capture. Therefore, acquiring images of the crack at different angles for measurement becomes essential. Currently, monocular camera acquisition technology primarily uses two-dimensional image processing for crack measurement; however, due to variations in the shooting angle, the calculated crack size often contains significant errors.

[0122] To verify the accuracy and reliability of the measurement method of this invention, this embodiment employs a multi-device simultaneous measurement technique, and the results are compared with those of monocular measurement, as detailed in Table 1. In this embodiment, two sets of representative images of the cracks to be detected were acquired using dual cameras, and named Group A and Group B, respectively. By comparing and analyzing the measurement results of different data acquisition methods, the advantage of the measurement method of this invention in terms of shooting angle can be more intuitively demonstrated, providing more accurate and reliable crack size data for practical applications.

[0123] Table 1 Comparison of data from different crack data acquisition methods

[0124]

[0125] As shown in Table 1, three different shooting angles were selected for image acquisition in each group of crack data collection, thus enabling a more comprehensive capture of the crack's morphology and size information from different perspectives. Subsequently, the width of the crack to be detected was measured according to the crack width measurement method of this invention, where the width specifically refers to the value at the widest point of the crack in the image.

[0126] To verify the beneficial effects of the crack width measurement method, this embodiment uses field measurements as standard reference values, and compares and analyzes the results and errors of the verification crack width measurement method. Furthermore, to more intuitively analyze the differences between this invention and existing verification measurement methods, and to more clearly demonstrate the changing trends of measurement errors, the data error results of different measurement methods are plotted as line graphs. Please refer to [link / reference] for details. Figure 4 .

[0127] This invention proposes a method and system for measuring crack width. Information detection equipment is deployed within a predetermined crack detection area to acquire initial crack detection data. Next, a detection data optimization model is constructed to optimize the initial data, resulting in more accurate optimized crack detection data. Subsequently, a detection data evaluation model is established, and based on relevant mathematical models and the optimized data, target data information for crack detection is further extracted. Based on the aforementioned target data information, the location information of the crack can be accurately analyzed. Finally, combining the crack width analysis model and the crack location information, the width of the crack to be measured is accurately measured. This invention achieves rapid extraction of crack location information and accurate measurement of crack width by constructing data optimization, evaluation, and measurement models.

[0128] Please see Figure 5 In an optional embodiment, to efficiently execute the crack width measurement method provided by the present invention, the present invention also provides a crack width measurement system. The crack width measurement system comprises an input device, a processor, an output device, and a memory interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to execute the specific steps of the relevant embodiments of the crack width measurement method provided by the present invention. The crack width measurement system of the present invention has a complete and stable structure, and can efficiently execute the crack width measurement method of the present invention, thereby improving the overall applicability and practical application capability of the present invention.

[0129] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A crack width measurement method characterized by, The method comprises the following steps: installing information monitoring equipment in a crack to-be-measured area, and obtaining crack detection initial data by using the information monitoring equipment; constructing a detection data optimization model to obtain crack detection optimized data by using the detection data optimization model; establishing a detection data evaluation model to obtain crack detection target data based on the detection data evaluation model and the crack detection optimized data; extracting crack detection position information according to the crack detection target data; measuring the width of the crack to-be-measured according to a crack width analysis model and the crack detection position information; the method for establishing the detection data evaluation model and obtaining the crack detection target data based on the detection data evaluation model and the crack detection optimized data comprises: constructing the detection data evaluation model by using the crack detection optimized data; obtaining a data evaluation result of the image to-be-detected by using the detection data evaluation model; the detection data evaluation model satisfies the following relationship: ， wherein, represents the evaluation result of the image data set to be detected, represents the total number of images to be detected, represents the initial data set of the image to be detected, represents the data set referenced by the detected image, represents the fluctuation coefficient of the data acquisition device, represents the external influence weight of the monitored object; the fluctuation coefficient of the data acquisition equipment is a numerical index of the data variation degree in the data acquisition process of the equipment, and the greater the fluctuation coefficient, the greater the data variation degree, and the smaller the fluctuation coefficient, the smaller the relative dispersion degree of the data; the external influence weight of the monitored object is the influence degree or proportion of external factors on the state or performance of the monitored object, and in the crack measurement process, the monitored object includes buildings and bridge structures which are affected by various external factors including wind, temperature, humidity, vibration and load; the method for constructing the detection data optimization model and obtaining the crack detection optimized data by using the detection data optimization model comprises: obtaining the pixel proportion of the image to-be-detected based on the crack detection initial data; establishing the detection data optimization model according to the pixel proportion; obtaining the crack detection optimized data by using the detection data optimization model; the detection data optimization model satisfies the following relationship: ， wherein, represents the image after optimization, represents the minimum output value of the converted image, represents the gray pixel proportion value of the image to be detected, represents the maximum output value of the converted image; the method for obtaining the crack detection target data based on the detection data evaluation model and the crack detection optimized data comprises: evaluating and analyzing the crack detection optimized data by using the detection data evaluation model, and obtaining a data evaluation result; obtaining a comparison result by comparing the data evaluation result to analyze the gap of the initial data, the comparison result includes accuracy, false detection rate, and missing detection rate, and iteratively optimizing the crack detection data based on the comparison result and the detection data optimization model; obtaining the crack detection target data by combining the data evaluation result and the detection data optimization model.

2. The crack width measurement method according to claim 1, characterized by, the method for installing the information monitoring equipment in the crack to-be-measured area and obtaining the crack detection initial data by using the information monitoring equipment comprises: installing the information monitoring equipment in the crack to-be-measured area, and the information monitoring equipment comprises a plurality of dynamic video cameras.

3. The crack width measurement method according to claim 1, characterized by, the method for extracting the crack detection position information according to the crack detection target data comprises: constructing a crack position correction model; obtaining the crack to-be-detected position information according to the crack position correction model and the crack detection target data.

4. The crack width measurement method according to claim 1, characterized by, the method for measuring the width of the crack to-be-measured according to the crack width analysis model and the crack detection position information comprises: constructing a crack width analysis model; the crack width analysis model satisfies the following relationship; ， wherein, represents a distance between any two detection points of the image to be detected, represents a first dimension of the first detection point of the image to be detected in a horizontal direction, represents a first dimension of the second detection point of the image to be detected in a horizontal direction, represents a second dimension of the first detection point of the image to be detected in a horizontal direction, represents a second dimension of the second detection point of the image to be detected in a horizontal direction, represents a third dimension of the first detection point of the image to be detected perpendicular to the horizontal direction of the image, represents a third dimension of the second detection point of the image to be detected perpendicular to the horizontal direction of the image.

5. A crack width measurement system characterized in that, The system comprises a processor, an input device, an output device and a memory, which are connected to each other, wherein the memory is configured to store a computer program, the computer program comprises program instructions, and the processor is configured to invoke the program instructions to execute the crack width measurement method according to any one of claims 1-4.

Images