A calibration method for a device used to remotely measure cracks

Abstract

The present invention discloses a calibration method for a device for remotely measuring cracks, which relates to the technical field of crack detection. The method includes: obtaining geometric parameter information of building cracks; determining at least one standard crack model based on the geometric parameter information of the building cracks; fabricating crack modules corresponding one-to-one to the respective standard crack models based on the respective standard crack models, and constructing a physical crack sample library based on the crack modules; selecting a target crack module, detecting the target crack module by the device for remotely measuring cracks to obtain first crack detection data; detecting the target crack module by a standard detection device to obtain second crack detection data, wherein the detection accuracy of the standard detection device is higher than that of the device for remotely measuring cracks; and determining the detection error of the device for remotely measuring cracks based on the first crack detection data and the second crack detection data. By establishing a standard model through the geometric parameters of the cracks, it is ensured that diverse crack types are covered, and the authenticity of the detection is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of crack detection, and particularly to a calibration method for a device for remotely measuring cracks. Background Art

[0002] Cracks on the surface of a building are a type of building disease, and are often affected by factors such as building settlement, fatigue stress, thermal expansion and contraction, and external loads, resulting in cracks on the surface of the structure.

[0003] Currently, crack detection technology has become the mainstream method due to its non-contact and high-efficiency characteristics. In the actual operation process, high-resolution imaging devices are usually used to obtain crack images, and subsequent spreading and analysis are carried out in the form of pictures. However, the resolution of the images obtained by current imaging devices can mostly only reach the level of 0.1 mm. But when facing the task of detecting micro-cracks, it is difficult to meet the precise measurement requirements of micro-cracks below 0.05 mm, there are significant errors in the identification of fine cracks, it is difficult to meet the measurement requirements of precision scenarios, and the geometric parameters of the measured cracks have large errors. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a calibration method for a device for remotely measuring cracks.

[0005] A calibration method for a device for remotely measuring cracks provided by the present invention, the calibration method for the device for remotely measuring cracks includes:

[0006] Obtain geometric parameter information of building cracks;

[0007] Based on the geometric parameter information of the building cracks, determine at least one standard crack model;

[0008] Based on each of the standard crack models, fabricate crack modules corresponding one-to-one to each of the standard crack models, and construct a physical crack sample library based on the crack modules;

[0009] Select a target crack module in the physical crack sample library, and detect the target crack module by a device for remotely measuring cracks to obtain first crack detection data;

[0010] Detect the target crack module by a standard detection device to obtain second crack detection data, and the detection accuracy of the standard detection device is higher than that of the device for remotely measuring cracks;

[0011] Based on the first crack detection data and the second crack detection data, determine the detection error of the device for remotely measuring cracks.

[0012] In some embodiments of the present invention, the step of obtaining geometric parameter information of building cracks includes:

[0013] The building is initially scanned by an image sensor carried by a drone to obtain an initial scanned image;

[0014] The initial scanned image is subjected to crack identification by an image recognition sensor carried by the drone to determine the crack area;

[0015] The crack area is scanned by a laser measuring instrument carried by the drone to obtain geometric parameter information of the building cracks.

[0016] In some embodiments of the present invention, the step of determining at least one standard crack model based on the geometric parameter information of the building cracks includes:

[0017] Data fitting processing is performed on the geometric parameter information of multiple building cracks to obtain at least one standard crack model, and one standard crack model is obtained by data fitting of the geometric parameter information of multiple building cracks.

[0018] In some embodiments of the present invention, the data fitting processing includes any one or any combination of linear regression fitting processing, polynomial fitting processing, spline interpolation fitting processing, surface fitting processing, and least squares fitting processing.

[0019] In some embodiments of the present invention, the step of detecting the target crack module by a remote crack measuring device to obtain first crack detection data includes;

[0020] The remote crack measuring device is used to detect the first crack width, first crack length, first crack depth, and first crack position coordinates of the target crack module;

[0021] Based on the first crack width, the first crack length, the first crack depth, and the first crack position coordinates, the first crack detection data is determined.

[0022] In some embodiments of the present invention, the step of detecting the target crack module by a standard detection device to obtain second crack detection data includes: detecting the second crack width, second crack length, second crack depth, and second crack position coordinates of the target crack module by the standard detection device;

[0023] Based on the second crack width, the second crack length, the second crack depth, and the second crack position coordinates, the second crack detection data is determined.

[0024] In some embodiments of the present invention, the crack module includes a first type of crack module, which is an integral structure. A complete crack is provided in the first crack module, and the first type of crack module is dimensionally matched with the standard detection device.

[0025] In some embodiments of the present invention, the crack module includes a second type of crack module, which includes a plurality of module units. A crack unit is provided on each of the module units. When the plurality of module units are in a spliced state, the plurality of crack units are spliced into a complete crack, and the module unit is dimensionally matched with the standard detection device.

[0026] In some embodiments of the present invention, the remote crack measurement device includes a drone, and the standard detection device includes a coordinate measuring machine and an industrial microscope.

[0027] In some embodiments of the present invention, the generation method of the crack module includes any one or a combination of any multiple of three-dimensional printing, precision machining, and three-axis computer numerical control machining.

[0028] The calibration method for the remote crack measurement device provided by the present invention has at least the following advantages:

[0029] The calibration method for the remote crack measurement device provided by the present invention establishes a standard crack model by collecting the geometric parameters of actual cracks, ensuring that the calibration covers diverse crack types, avoiding the limitations of a single scenario, transforming the mathematical model into a physical crack module, constructing a real environment for repeatable testing, and calibrating the errors of the remote crack measurement device based on the detection data of the high-precision standard detection device to ensure the authority of the comparison. By comparing the first crack detection data and the second crack detection data, the detection error of the remote crack measurement device can be determined, ensuring the authenticity of the detection error and improving the applicability, accuracy, and precision of the error detection method. Description of the Drawings

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0031] Figure 1 It is a flowchart of the calibration method for the remote crack measurement device provided by an exemplary embodiment of the present invention. Detailed Embodiments

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the protection scope of the present invention.

[0033] Referring to "embodiments" herein means that the specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present invention. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0034] Cracks on the surface of a building are a type of building disease and are often affected by building settlement, fatigue stress, thermal expansion and contraction, and external loads, resulting in cracks on the surface of the structure.

[0035] At present, crack detection technology has become the mainstream method due to its non-contact and high-efficiency characteristics. In the actual operation process, high-resolution imaging equipment is usually used to obtain crack images and subsequent spreading and analysis are carried out in the form of pictures. The resolution of the images obtained by current imaging equipment can mostly only reach the level of 0.1 mm. However, when facing the task of detecting micro-cracks, it is difficult to meet the accurate measurement requirements of micro-cracks below 0.05 mm, there are significant errors in the identification of fine cracks, it is difficult to meet the measurement requirements of precision scenarios, and the geometric parameters of the measured cracks have large errors.

[0036] Next, in conjunction with Figure 1 a further description of the embodiments of the present invention will be given.

[0037] To solve the above technical problems, the present invention provides a calibration method for a remote crack measurement device. By collecting the geometric parameters of actual cracks, a standard model is established to ensure that the calibration covers diverse crack types, avoiding the limitations of a single scenario. The mathematical model is transformed into a physical crack module to construct a repeatable test real environment. Based on the detection data of a high-precision standard device, the errors of the standard detection device are calibrated to ensure the authority of the comparison. By comparing the first crack detection data and the second crack detection data, the detection error of the remote crack measurement device can be determined, ensuring the authenticity of the detection error and improving the applicability and accuracy of the error detection method.

[0038] An exemplary embodiment of the present invention provides a calibration method for a remote crack measurement device, as Figure 1 shown, the calibration method for a remote crack measurement device includes:

[0039] S100. Obtain the geometric parameter information of building cracks.

[0040] In step S100, a drone equipped with an image sensor can be used to conduct a preliminary scan of the building to obtain a preliminary scan image. Then, a crack recognition sensor is used to identify cracks in the preliminary scan image to determine the crack area. Furthermore, a laser measuring instrument is used to scan the crack area to measure the actual building cracks, and a true crack sample with geometric parameter data such as the width, depth, length, orientation, and position information of various cracks is obtained. A three-dimensional digital model of the crack is generated through point cloud processing software, and key geometric parameter data, such as the width, depth, length, and position information of the crack, are extracted to form the geometric parameter information of the building crack. In this way, the geometric parameter information can cover the geometric parameters of the vast majority of cracks, ensuring the integrity of the acquisition of crack geometric parameter information. Based on actual building data modeling, it avoids the deviation caused by theoretical assumptions, combines the surface and internal parameters of the crack, covers the full-dimensional characteristics of the crack, and constructs a real test environment.

[0041] S200. Determine at least one standard crack model based on the geometric parameter information of the building crack.

[0042] In step S200, the geometric parameter information of the building crack can be processed through data fitting to establish a more comprehensive and accurate standard crack model. In addition, big data and AI simulation technologies can be used to process the standard crack model to form a standard two-dimensional or three-dimensional crack model. By integrating crack data of different shapes and sizes, the constructed model can adapt to a wider range of building crack scenarios, avoid the overfitting problem of single crack fitting, and the unified model based on multi-crack fitting can calibrate the detection ability of remote crack measurement devices for different types of cracks at one time.

[0043] S300. Based on each standard crack model, make crack modules corresponding to each standard crack model one by one, and construct a physical crack sample library based on the crack modules.

[0044] In step S300, the standard crack models are imported into the manufacturing software. The processing technology is selected according to material properties (such as concrete, metal, composite materials) and crack complexity, such as 3D printing, three-axis computer numerical control (CNC), precision casting, to generate crack modules corresponding one by one to the standard crack models, and then a physical crack sample library is constructed based on the crack modules. The physical crack sample library can cover all types of cracks and adapt to different scenarios such as bridges, tunnels, and ancient buildings. This enables the crack modules to simulate different crack scenarios, improving the authenticity and accuracy of the calibration of remote crack measurement devices. By collecting the geometric parameters of actual cracks to establish a standard model, it is ensured that the calibration covers diverse crack types, avoiding the limitations of a single scenario. The mathematical model is transformed into an entity crack module to construct a repeatable test real environment. Based on the detection data of high-precision standard devices, the errors of remote crack measurement devices are calibrated to ensure the authority of the comparison.

[0045] S400. Select a target crack module from the physical crack sample library, and detect the target crack module through a remote crack measurement device to obtain first crack detection data.

[0046] In step S400, a target crack module is selected from the physical crack sample library, where the target crack module is suitable for showing the type of crack to be actually measured. The target crack module is detected through a remote crack measurement device. For example, the crack area is scanned by a drone equipped with an image sensor, an image recognition sensor, and a laser rangefinder to obtain the geometric parameter information of the crack in the target crack module. By selecting a target crack module that highly matches the type of crack in the building to be measured, the actual detection environment is reproduced to verify the performance of the remote crack measurement device in a real scenario.

[0047] S500. Detect the target crack module through a standard detection device to obtain second crack detection data, and the detection accuracy of the standard detection device is higher than that of the remote crack measurement device.

[0048] In step S500, the standard detection device can use an industrial microscope or an industrial coordinate measuring machine to assign values to the geometric parameter information of the crack, and its accuracy can reach 0.003 mm. It can also be cross-verified through multi-sensor fusion. For example, an industrial microscope and an industrial coordinate measuring machine are used for measurement simultaneously to ensure the absolute credibility of the reference data. Different sensors have different measurement principles and advantages. By comparing and analyzing their measurement results, they can confirm each other and eliminate errors and interference factors in the measurement process. The geometric parameter information of the crack in the target crack module is calibrated through a high-precision standard detection device to obtain second crack detection data. The second crack detection data, as high-precision standard data, provides standard data for the performance verification of the remote crack measurement device, and further makes the error calculation more reliable.

[0049] S600. Determine the detection error of the remote crack measurement device based on the first crack detection data and the second crack detection data.

[0050] The first crack detection data of the target crack module is directly measured by the remote crack measurement device. Then, the target crack module is detected by a standard detection device to obtain the second crack detection value. The second crack detection data is used as the reference value of the benchmark. By calculating the absolute error and relative error between the first crack detection data and the second crack detection data, the maximum allowable error is set. The root cause of hardware or software problems is identified through error decomposition. For example, an AI model (such as a regression model) is trained using the error data to automatically compensate for environmental interference (such as the metal expansion error caused by temperature). Furthermore, the accuracy and precision values of the remote crack measurement device are improved. By comparing the first crack detection data measured by the remote crack measurement device with the second crack detection data measured by the standard detection device, the accuracy and reliability of the remote crack measurement device in measuring various parameters can be intuitively evaluated. For example, if there is a large deviation between the crack width measured by the remote crack measurement device and the second crack detection data, it indicates that there are performance problems with the device in width measurement and calibration or optimization is required.

[0051] In some embodiments of the present invention, step S100 of obtaining the geometric parameter information of the building crack includes:

[0052] S101. Conduct a preliminary scan of the building through an image sensor carried by a drone to obtain a preliminary scan image.

[0053] Based on the building CAD model or satellite map, a three-dimensional grid flight path is generated to ensure coverage of the building to be detected. The drone automatically flies according to the preset path, reducing redundant waypoints. The image sensor continuously captures images at a preset frequency and transmits them back to the ground station in real time. The ground station software checks the image quality in real time and automatically marks the areas that need to be reshot. In addition, parameters such as the focal length and exposure time of the image sensor need to be adjusted to ensure clear and accurate images can be obtained under different lighting conditions. In this way, the acquisition efficiency can be improved, manual detection blind spots can be avoided, a healthy thermal map of the building surface can be quickly generated, and potential hazard areas can be marked. The efficiency of building crack detection is improved. By real-time image quality inspection, invalid data is prevented from entering the subsequent analysis process, reducing waste of computing resources.

[0054] S102. Identify cracks in the preliminary scan image through an image recognition sensor carried by the drone to determine the crack area.

[0055] Frame-by-frame analysis is performed on the preliminary scanned images in step S101 through an image recognition sensor to identify the crack pixel regions, and a binary mask image is output. Combining the three-dimensional point cloud coordinates, the cracks are mapped to the real building coordinate system to obtain the crack coordinates. Through pixel-point cloud fusion, the accuracy of crack positioning is improved, the crack regions are automatically marked, reducing the workload of manual annotation. The image recognition sensor is trained and optimized to improve its accuracy and robustness in crack recognition.

[0056] S103. Scan the crack region with the laser measuring instrument carried by the UAV to obtain the geometric parameter information of the building cracks.

[0057] Based on the crack coordinates in step S102, plan the UAV's flight path close to the building, set the scan line density, enable the laser measuring instrument to scan the building, collect the three-dimensional point cloud of the cracks, and synchronously record the environmental parameters for data compensation. For the various data collected, use image processing software and data analysis algorithms to obtain the geometric parameter information of the building cracks. Among them, for the image data, the contour boundaries of the cracks can be identified through edge detection algorithms (such as the Canny edge detection algorithm), and basic parameters such as the length, width, and orientation angle of the cracks can be calculated. Combining the three-dimensional laser scanning data, using point cloud processing technology, accurately measure parameters such as the depth, surface roughness, and bending degree of the cracks in three-dimensional space. Exemplarily, for example, when measuring the crack length, the center line of the crack can be obtained by fitting the points on the crack contour, and then the length of the center line is calculated; when measuring the crack width, multiple positions on the crack cross-section are selected for measurement, and the average value is taken as the crack width value. Organize and classify the extracted various geometric parameter information to form a detailed geometric parameter database of building cracks, providing a data basis for subsequent crack morphology screening and data fitting processing. In addition, the laser measuring instrument also needs to be calibrated to eliminate measurement errors and ensure the reliability of the measurement data.

[0058] In some embodiments of the present invention, the step S200 of determining at least one standard crack model based on the geometric parameter information of building cracks includes: performing data fitting processing on the geometric parameter information of multiple building cracks to obtain at least one standard crack model, where one standard crack model is obtained by performing data fitting on the geometric parameter information of multiple building cracks. Among them, the crack morphologies collected are initially screened, and the cracks with the same morphology are taken as the same type of cracks for data fitting processing. Among them, the cosine similarity algorithm can be used to calculate the similarity of the crack orientation angles, and the Euclidean distance is used to calculate the similarity of numerical parameters such as crack length and width. For each geometric parameter, a corresponding weight is set, and a reasonable distribution is made according to the influence degree of each parameter on the crack morphology. The similarity values of each parameter are weighted and summed to obtain the comprehensive similarity score of the crack morphology. A similarity threshold is set. When the comprehensive similarity score of two cracks is higher than this threshold, it is considered that the two cracks have the same morphology and can be classified into the same type of cracks.

[0059] Then, data fitting processing is performed on each type of crack respectively. Specifically, based on the geometric parameter information of the original building cracks, the multiple geometric parameter information is divided into multiple first initial data sets, and the multiple first initial data sets correspond one by one to multiple preset crack regions; the first initial data set corresponding to the target crack region is determined as the first target data set, and data fitting processing is performed on the geometric parameters in the first target data set to obtain at least one standard crack model. Specifically, a suitable mathematical model and fitting algorithm are used to perform data fitting processing on the geometric parameters in the first target data set. For the crack orientation and morphology, common fitting methods include polynomial fitting, spline curve fitting, etc. For example, for the change in the crack orientation, if it shows a relatively smooth curve feature, the cubic spline curve fitting method can be used. Through the known crack orientation point coordinates, a smooth curve that can accurately describe the crack orientation is fitted. When fitting parameters such as crack length and width, considering that there may be a certain distribution law for crack parameters, a probability distribution fitting method is used. Assuming that the crack length follows a normal distribution, by statistically analyzing the crack length data in the first target data set, its mean and standard deviation are calculated, and then the parameters of the normal distribution are determined to obtain the probability distribution model of the crack length.

[0060] Among them, the method of data fitting processing can adopt any one or any combination of linear regression fitting processing, polynomial fitting processing, spline interpolation fitting processing, surface fitting processing, and least squares fitting processing as the fitting method for data fitting processing of the geometric parameter information of multiple building cracks. It can utilize the advantages of each fitting processing method to fit a large amount of original data, enabling the standard crack model to have a strong fitting effect and meet different scenario requirements. It can be applied to different crack detections, enabling the standard crack model to calibrate and cover diverse crack types, avoiding the limitations of a single scenario. By comprehensively applying multiple fitting methods, comprehensive data fitting processing is carried out on the geometric parameters of building cracks, and finally at least one standard crack model that can accurately describe the typical crack morphology of the target crack area is obtained. This standard crack model not only includes the main geometric feature parameters of the crack but also reflects the morphological distribution law of the crack in the target area, providing an important reference basis for subsequent building structure safety assessment, crack cause analysis, repair plan design, and calibration of remote crack measurement devices. In addition, the mathematical model is transformed into an entity crack module to construct a repeatable test real environment. Based on the detection data of high-precision standard devices, the errors of remote crack measurement devices are calibrated to ensure the authority of the comparison.

[0061] In some embodiments of the present invention, step S400, the step of obtaining the first crack detection data by detecting the target crack module with a remote crack measurement device includes:

[0062] Detecting the first crack width, first crack length, first crack depth, and first crack position coordinates of the target crack module with a remote crack measurement device;

[0063] Based on the first crack width, first crack length, first crack depth, and first crack position coordinates, determining the first crack detection data.

[0064] Select the target crack module in the physical crack sample library. When selecting the target crack module, it is necessary to closely combine the actual crack type to be measured. Among them, the target crack module is suitable for displaying the actual crack type to be measured. The target crack module is detected by remote crack measurement equipment. For example, the crack area is scanned by an image sensor, an image recognition sensor and a laser measuring instrument carried by a drone. The image sensor carried on the drone has high-resolution imaging capabilities and can clearly capture information such as the surface texture and morphological details of the crack; the image recognition sensor uses an image recognition algorithm to automatically identify the crack area and extract the key features of the crack, such as the crack edge contour; the laser measuring instrument is used to accurately measure the geometric parameters of the crack, such as width and depth. Among them, the first crack detection data includes the first crack width, first crack length, first crack depth and first crack position coordinates of the target crack module. By selecting a target crack module that is highly matched with the type of building crack to be measured, the actual detection environment is reproduced to verify the performance of the remote crack measurement equipment in a real scene. By using remote crack measurement equipment to measure the first crack width, length, depth and position coordinates, and comparing them with the standard values measured by high-precision standard equipment, the error size of the remote crack measurement equipment in measuring various parameters can be accurately calculated, including absolute error and relative error. For example, if the first crack width measured by the laser measuring instrument deviates greatly from the standard value, it can be clear that the equipment has accuracy problems in width measurement, so that the equipment can be calibrated and optimized in a targeted manner. Based on the first crack detection data, the image recognition algorithm can be optimized. When extracting key features such as the edge contour of the crack, the image recognition sensor may have inaccurate recognition or omission. By analyzing a large amount of first crack detection data, the performance of the algorithm under different crack shapes and lighting conditions can be understood, and the causes of recognition errors can be found, and then the parameters of the algorithm can be adjusted or the algorithm model can be improved to improve the accuracy and robustness of crack recognition. For example, if the algorithm is not effective in identifying small cracks, the algorithm performance can be improved by increasing data training samples and optimizing feature extraction methods. Then, the crack geometric parameter information of the target crack module is calibrated using high-precision standard detection equipment to provide standard data for the remote crack measurement equipment, thereby making the error calculation more reliable.

[0065] In some embodiments of the present invention, step S500, detecting the target crack module by using a standard detection device to obtain second crack detection data, includes:

[0066] Detect the second crack width, the second crack length, the second crack depth and the second crack position coordinates of the target crack module by using standard detection equipment;

[0067] Determine the second crack detection data based on the second crack width, second crack length, second crack depth, and second crack position coordinates.

[0068] The standard detection equipment can use an industrial microscope and an industrial coordinate measuring instrument to assign values to the geometric parameter information of the crack, and its accuracy can reach 0.003 mm. To improve the accuracy and reliability of the crack geometric parameter measurement, multi-sensor fusion cross-validation can also be used. For example, the industrial microscope and the industrial coordinate measuring instrument can measure simultaneously to ensure the absolute credibility of the reference data. Different sensors have different measurement principles and advantages. By comparing and analyzing their measurement results, they can corroborate each other and eliminate errors and interference factors in the measurement process. Calibrate the crack geometric parameter information of the target crack module using high-precision standard detection equipment to obtain the second crack detection data, which provides standard data for the remote crack measurement equipment, thereby making the error calculation more reliable. Exemplarily, when performing multi-sensor fusion cross-validation, first use the industrial microscope to preliminarily observe and measure the crack to obtain the microscopic image and partial geometric parameter information of the crack. Then, use the industrial coordinate measuring instrument to measure at the same position of the crack to obtain the three-dimensional coordinate data and more accurate geometric parameters of the crack. Compare the measurement results of the two sensors. If the error between them is within the allowable range, the measurement result can be considered reliable; if the error is large, the reason needs to be further analyzed, which may be caused by improper measurement methods, sensor failures, or crack surface characteristics, etc., and then corresponding measures are taken for correction and re-measurement.

[0069] The second crack detection data, as high-precision standard data, provides an accurate reference for the performance verification of the remote crack measurement equipment. By comparing the first crack detection data measured by the remote crack measurement equipment with the second crack detection data measured by the standard detection equipment, the accuracy and reliability of the remote crack measurement equipment in various parameter measurements can be intuitively evaluated. For example, if there is a large deviation between the crack width measured by the remote crack measurement equipment and the second crack detection data, it indicates that the equipment has performance problems in width measurement and needs to be calibrated or optimized.

[0070] In some embodiments of the present invention, the crack module includes a first type of crack module. The first type of crack module is an integral structure. A complete crack is provided in the first crack module, and the first type of crack module is dimensionally matched with the standard detection equipment. In this embodiment, the module is a single-piece integral structure, the crack is complete and continuous without breakpoints, and the dimensions are strictly matched with the standard detection equipment. It is formed in one piece by 3D printing, CNC machining or precision casting to ensure a high degree of consistency between the geometric parameters of the crack (such as width, depth, length, etc.) and the actual situation. The first crack module can be directly placed on the workbench of the standard detection equipment without assembly and debugging, reducing human operation errors. At the same time, there are no moving parts or seams, and it is not easy to deform or wear during long-term use. This makes the data obtained by the remote crack measurement device highly coherent and complete. For example, when using a laser measuring instrument to measure the depth of a crack, the complete crack can ensure the continuous propagation of the measuring light in the crack, thereby obtaining accurate depth data.

[0071] In some embodiments of the present invention, the crack module includes a second type of crack module. The second type of crack module includes a plurality of module units, and a crack unit is provided on each module unit. When the plurality of module units are in a spliced state, the plurality of crack units are spliced into a complete crack, and the module unit is dimensionally matched with the standard detection equipment. In this embodiment, it is composed of several independent module units (the size is, for example, 50mm×50mm), and each unit contains a local crack. After splicing, a complete crack is formed (such as a crack with a total length of 1m). Each unit is processed separately, and the interfaces are designed as tenon and mortise, magnetic adsorption or precision guide rails to ensure the splicing accuracy. By increasing or decreasing the module units, the crack length (for example, expanding the crack length from 0.5m to 3m) and shape (such as straight line, broken line, mesh) can be dynamically adjusted to adapt to different detection scenarios. Different types of crack units (such as a unit with a width of 0.1mm + a unit with a depth of 5mm) can be combined to simulate composite cracks in real buildings.

[0072] In some embodiments of the present invention, the remote crack measurement device includes a drone, and the standard detection device includes a coordinate measuring machine and an industrial microscope. In this embodiment, the remote crack measurement device can use a drone equipped with an image sensor, an image recognition sensor, and a laser scanner to scan the crack area to obtain the geometric parameter information of the building crack. The drone can fly to high-altitude, narrow, or dangerous areas, quickly scan a large surface area through the image sensor (RGB camera), and the laser scanner synchronously obtains the three-dimensional point cloud data of the crack to improve the data acquisition efficiency. The measurement accuracy of the geometric parameters of the target crack module by the coordinate measuring machine can reach ±0.001 mm. The industrial microscope magnifies 1000 times to observe the roughness of the crack edge and the bifurcation of microcracks to assist in judging the cause of the crack. The results measured by the standard detection device are used as the standard to calibrate the drone data, and the system error of the drone can be accurately calculated, and the sensor parameters can be corrected accordingly. Establish a traceable metrological traceability chain, and the drone measurement data is traced back to national / international standards (such as ISO certification) through the standard detection device to ensure the legal effect and cross-platform comparability of the detection results. Through the laboratory standard detection device - drone calibration - on-site detection - laboratory standard detection device re-measurement, a technical closed-loop is formed to continuously verify the performance stability of the drone.

[0073] In some embodiments of the present invention, the generation method of the crack module includes any one or any combination of three-dimensional printing, precision machining, and three-axis computer numerical control machining. It can be understood that any one or any combination of three-dimensional printing, precision machining, and three-axis computer numerical control machining can be used when generating the crack module. Among them, three-dimensional printing can generate complex shapes such as internal cavities and bifurcated cracks that are difficult to achieve by traditional machining. By replacing the printing material, the crack characteristics of different substrates such as concrete, metal, and plastic can be simulated. Precision machining (traditional machining) directly uses engineering materials for processing and can retain the original mechanical and thermal properties. Machining can achieve a dimensional accuracy of ±0.01 mm, which is suitable for calibrating high-sensitivity detection equipment and suitable for producing a small number of high-precision standard crack modules. Three-axis computer numerical control (CNC) machining can automatically process dozens of modules per hour, which is suitable for large-scale calibration requirements. Three-axis linkage can process medium-complexity shapes such as inclined cracks and wavy edges to ensure that the geometric parameters of all modules are highly unified.

[0074] In another embodiment, after step S300, the step of fabricating crack modules corresponding to each standard crack model based on each standard crack model further includes: performing process error detection on each generated crack module to determine the dimensional error between the crack data of the crack module caused by the processing technology of the crack module and the standard crack model; detecting the crack modules with dimensional errors less than a preset error through a standard detection device. In this way, by screening out the crack templates with errors greater than the preset error, the transfer of processing errors to subsequent detection links is avoided, the detection cost can be saved, and the detection efficiency can be improved.

[0075] It can be understood that when fabricating crack modules corresponding to each standard crack model through the standard crack model, due to the limitations of the processing technology, there may be a certain dimensional error between the set of parameter information of the cracks in the generated crack modules and the standard crack model. To ensure the accuracy and reliability of subsequent detection work and avoid interference with the detection results caused by the processing errors of the crack modules themselves, process error detection can be performed on each generated crack module to determine the error range of each crack module through the process error detection. Exemplarily, in the actual operation process, a high-precision three-dimensional scanning and measuring device can be used to perform a full-range scan of the crack module. This device can capture every fine feature on the surface of the crack module with extremely high resolution and generate detailed three-dimensional point cloud data. Subsequently, these point cloud data are accurately compared with the three-dimensional data of the pre-constructed standard crack model. Through existing data processing software, the deviation values of the crack module from the standard crack model in each key dimensional parameter are calculated, thereby obtaining comprehensive dimensional error data. If the dimensional error is greater than the preset error, it indicates that the reduction degree of the geometric parameter information in the standard crack model by the crack module is insufficient, which may further affect the subsequent calibration results. Based on the error detection results, the crack modules with dimensional errors less than the preset error are screened out as qualified modules meeting the detection requirements; while the crack modules with dimensional errors greater than the preset error are regarded as unqualified modules.

[0076] By performing process error detection on each generated crack module and screening out the crack modules meeting the requirements, the crack templates with large processing errors can be effectively removed at an early stage, avoiding the transfer of processing errors to subsequent detection links. This helps to save the additional detection cost caused by detecting unqualified modules and can also significantly improve the overall detection efficiency. By detecting the crack modules meeting the requirements and comparing the first crack detection data and the second crack detection data, the detection error of the remote measurement crack device can be determined, ensuring the authenticity of the detection error, enhancing the applicability and accuracy of the error detection method, and ensuring that the finally obtained detection results can truly and accurately reflect the performance characteristics of the crack module.

[0077] The content described above can be implemented alone or in various combinations, and these variations are all within the scope of protection of the present invention.

[0078] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A calibration method for a device used to remotely measure cracks, characterized in that, The calibration method for the device used to remotely measure cracks includes: Obtaining geometric parameter information of building cracks; Based on the geometric parameter information of the building cracks, determining at least one standard crack model. Among them, the step of determining at least one standard crack model based on the geometric parameter information of the building cracks includes: performing data fitting processing on the geometric parameter information of multiple building cracks to obtain at least one standard crack model, and one standard crack model is obtained by performing data fitting on the geometric parameter information of multiple building cracks; Based on each of the standard crack models, manufacturing crack modules corresponding to each of the standard crack models one by one, and constructing a physical crack sample library based on the crack modules; Selecting a target crack module in the physical crack sample library, and detecting the target crack module by a device for remotely measuring cracks to obtain first crack detection data. Among them, the method of detecting the target crack module by a device for remotely measuring cracks to obtain first crack detection data includes: obtaining an image of a target area by an image sensor carried by a drone; automatically identifying a crack area by scanning the image of the target area with an image recognition sensor carried on the drone; detecting the target crack module by a laser measuring instrument carried by the drone to obtain first crack detection data, and the first crack detection data includes the first crack width, the first crack length, the first crack depth and the first crack position coordinates of the target crack module; Detecting the target crack module by a standard detection device to obtain second crack detection data, and the detection accuracy of the standard detection device is higher than that of the device for remotely measuring cracks; Based on the first crack detection data and the second crack detection data, determining the detection error of the device for remotely measuring cracks.

2. The calibration method for the remote crack measurement device according to claim 1, characterized in that The step of obtaining the geometric parameter information of building cracks includes: Performing a preliminary scan of the building by an image sensor carried by a drone to obtain a preliminary scan image; Identifying cracks in the preliminary scan image by the image recognition sensor carried by the drone to determine the crack area; Scanning the crack area by a laser measuring instrument carried by the drone to obtain the geometric parameter information of the building cracks.

3. The calibration method for the remote crack measurement device according to claim 1, characterized in that, The data fitting processing includes any one or any combination of linear regression fitting processing, polynomial fitting processing, spline interpolation fitting processing, surface fitting processing, and least squares fitting processing.

4. The calibration method for the remote crack measurement device according to claim 1, characterized in that, The step of detecting the target crack module by a device for remotely measuring cracks to obtain first crack detection data includes; Detecting the first crack width, the first crack length, the first crack depth and the first crack position coordinates of the target crack module by the device for remotely measuring cracks; Based on the first crack width, the first crack length, the first crack depth and the first crack position coordinates, determining the first crack detection data.

5. The calibration method for the remote crack measurement device according to claim 1, characterized in that, The step of detecting the target crack module by the standard detection device to obtain the second crack detection data includes: detecting the second crack width, the second crack length, the second crack depth and the second crack position coordinates of the target crack module by the standard detection device; Based on the second crack width, the second crack length, the second crack depth and the second crack position coordinates, determine the second crack detection data.

6. The calibration method for a remote crack measurement device according to any one of claims 1 to 5, characterized in that The crack module includes a first type of crack module, the first type of crack module is an integral structure, a complete crack is provided in the first crack module, and the first type of crack module is dimensionally matched with the standard detection device.

7. The calibration method for a remote crack measurement device according to claim 6, characterized in that, The crack module includes a second type of crack module, the second type of crack module includes a plurality of module units, a crack unit is provided on each of the module units, and when the plurality of module units are spliced, the plurality of crack units are spliced into a complete crack, and the module unit is dimensionally matched with the standard detection device.

8. The calibration method for a device for remotely measuring cracks according to any one of claims 1 to 5, characterized in that, The remote crack measurement device includes a drone, and the standard detection device includes a coordinate measuring machine and an industrial microscope.

9. The calibration method for the remote crack measurement device according to any one of claims 1 to 5, characterized in that, The manufacturing method of the crack module includes any one or any combination of three-dimensional printing, precision machining, and three-axis computer numerical control machining.

Images