Ray-based concrete pole defect detection method and equipment

Abstract

The invention discloses a ray-based concrete pole defect detection method and equipment, relates to the technical field of fault detection, and solves the problems that in the prior art, the integrity of a training data set is difficult to guarantee, so that the defect identification precision of a concrete pole is not high, and the defect identification is not comprehensive enough. The method comprises the following steps: acquiring dark field and bright field ray images of the same concrete pole area, pairing to form a first data set, taking the ray image of the first data set as a background, injecting a physical defect model generated by simulation, synthesizing a defect sample, training a fault identification model by using the defect sample, and identifying the fault by using the trained fault identification model. And identifying whether the ray image of the concrete pole has defects or not. According to the invention, portable detection of hidden dangers in the ground part of the concrete pole can be realized, the operation and maintenance efficiency is greatly improved, manpower and material resource expenditures are saved, breakage of main reinforcements of the concrete pole can be found in time, fault expansion is avoided, the maintenance cost is reduced, and remarkable direct economic benefits are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fault detection, in particular to a concrete pole defect detection method and equipment based on rays. BACKGROUND

[0002] As an important supporting structure of power transmission and distribution lines, concrete poles are widely used in urban and rural power transmission and distribution networks and play an important role in supporting lines and ensuring the safe and stable operation of power systems. Due to their high strength, corrosion resistance, and long service life, concrete poles have gradually replaced wooden and metal poles and become an indispensable infrastructure in power transmission and distribution lines. Especially in rural power grid reconstruction, new energy grid connection and urban power distribution network, the application range of concrete poles is continuously expanding, and their fracture risk directly threatens the safe and stable operation of power grids, personal and public safety. The fracture sites of concrete poles are often concentrated in the visible part above the ground, which is exposed to complex environments and mechanical loads for a long time, and is limited by the coverage and accuracy of existing detection techniques. Hidden dangers are difficult to be found in time. From the ground pole body to the wire connection part, the deterioration of every detail can become the starting point of fracture. For example, the weathering and carbonation process of the concrete surface appears as color whitening or powdering in vision, but deep carbonation will reduce the alkalinity of the concrete, causing internal steel corrosion and expansion, and eventually causing longitudinal cracks. These cracks may only appear as hairline cracks in the early stage, and if there is no high-precision detection means (such as unmanned aerial vehicle high-definition camera or laser scanning), they are easily overlooked by traditional manual inspection. In areas with frequent typhoons, dynamic wind loads on the pole can exacerbate the expansion of existing micro-cracks, especially at the pole body and cross arm connection, cracks may extend radially around the bolt hole due to stress concentration effect, eventually leading to flange fracture, in addition, internal steel corrosion and fracture of concrete poles can also greatly reduce the mechanical properties of the poles. Such damage is often missed in conventional ground visual inspection due to angle limitations and requires close-range inspection using climbing or lifting equipment, but due to the cost of operation and maintenance, such deep detection cannot be implemented frequently.

[0003] However, the current recognition of concrete poles usually collects images of the surface of the concrete pole to train the fault model, which cannot identify whether there are defects inside the concrete pole, and most of the current defect recognition uses existing defect images with corresponding fault labels to train the fault model, which requires a large number of real defect images for training. However, not every kind of defect occurs frequently, making it difficult to ensure the integrity of the training data set, and thus the defect recognition accuracy of the concrete pole is not high and not comprehensive.

[0004] Therefore, a concrete pole defect detection method and equipment based on rays are needed. SUMMARY

[0005] In order to solve the problem of low defect identification accuracy of the concrete pole in the prior art due to the difficulty in ensuring the integrity of the training data set, the application provides a concrete pole defect detection method and equipment based on rays, which can utilize the attenuation characteristics of rays (X-rays, gamma rays or neutron rays) during propagation. When the rays with uniform intensity are injected into the detected part from one side, the intensity of the rays after penetrating the detected part will be uneven due to the different attenuation characteristics of the defects and the base material of the detected part. The intensity of the rays after penetrating the detected part can be detected on the opposite side by methods such as film photography and direct observation of the fluorescent screen, so as to determine whether there are defects on the surface or inside of the detected part. The ray detection method has significant advantages in detection accuracy and is particularly suitable for large-scale detection, which can effectively evaluate the internal structure of the concrete pole. The specific technical solutions are as follows: A concrete pole defect detection method based on rays, comprising the following steps: Install a horizontal multi-angle tool including a ray source and an imaging plate, and install the tool and the related power device carried by the tool on the concrete pole, keeping the imaging plate and the ray source in a uniform horizontal angle; Remote start the ray machine, and perform ray penetration on the position to be detected, and read the ray image data from the imaging plate; Move the tool up and down and rotate the tool horizontally by the related power device carried by the tool to take multiple-angle photos of the possible defect positions of the concrete pole, and obtain multiple ray image data; Collect dark field and bright field ray images of the same concrete pole area, pair them to form a first data set, and use the ray images of the first data set as the background to inject a physically simulated defect model to synthesize a defect sample; Train a fault identification model using the defect sample, and use the trained fault identification model to identify whether the ray image of the concrete pole has defects.

[0006] Preferably, the specific steps of synthesizing the defect sample are as follows: Real ray image acquisition: acquire a DR image of a defect-free concrete pole as a background base; Defect physical simulation: at least simulate steel bar fracture, concrete hollowing and steel bar bending; Ray attenuation model fusion: use the DR image of the defect-free concrete pole as the background base, superimpose the simulated defects on the background and the ground, and label the fused ray image with the corresponding fault label.

[0007] Preferably, the defect physical simulation is specifically defect parameterization modeling, and the defect parameterization modeling is specifically simulating each parameterized defect of the defect, and the parameters at least include: fracture position, width, cross-section angle, hollowing, cavity volume, depth and shape.

[0008] Preferably, the process of synthesizing the defect sample further comprises noise injection, specifically, adding noise to the fused radiographic image after the radiographic attenuation model fusion and the fault labeling.

[0009] Preferably, the fault label at least includes fault type, fault depth, width, and depth ratio.

[0010] Preferably, after obtaining multiple radiographic image data, first, the feature information in the image is extracted, including steel reinforcement contour, edge, intersection, then, the three-point positioning method is used to obtain the initial estimated coordinates of these feature points in three-dimensional space, and the corresponding depth value is obtained by referring to the scene setting or physical size. Then, geometric transformation calculation is performed based on the single-point perspective projection model, and the model expression is: Wherein: , is the horizontal and vertical coordinates of the steel reinforcement in space; is the distance from the point to the center of the camera projection; is the focal length of the camera or perspective projection model; , is the corresponding projection point coordinate on the image.

[0011] After calculating the ideal projection coordinates of all key points through the formula, pixel mapping is performed combined with the image registration relationship to complete the first round of geometric correction.

[0012] Preferably, in the process of constructing and training the fault recognition model, at least the geometric features of the defect, the gray scale features of the defect, and the similarity between the image and the defect template in the image library are extracted during feature extraction, wherein the geometric features of the defect at least include the length of the steel reinforcement fracture and the area of the concrete hollow, and the gray scale features of the defect at least include the average gray scale value of the defect area.

[0013] A radiographic concrete pole defect detection device applied to the method described above, comprising: A horizontal multi-angle tooling, including a radiographic source and an imaging plate, wherein the radiographic source is used to emit radiographic rays to penetrate the concrete pole, and is commonly accompanied by a radiographic image receiver after penetrating the concrete pole, the radiographic source and the imaging plate are arranged on the same horizontal line, and the radiographic source and the imaging plate are arranged at both ends of the concrete pole ​The relevant power unit is connected to the horizontal multi-angle tooling and is used to carry the tooling to complete the X-ray image acquisition operation. The relevant power unit includes at least one of a fixed mounting bracket and a drone. The fixed mounting bracket is sleeved on the concrete pole and can rotate around the concrete pole. The X-ray source and the imaging plate are respectively connected to the fixed mounting bracket. If the relevant power unit is a drone, the drone is at least connected to the X-ray source, and the imaging plate is driven to rise and fall to the same horizontal line as the X-ray source by connecting to either the telescopic pole or the drone.

[0014] A computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the X-ray-based concrete pole defect detection method as described above.

[0015] A processor for running a program, wherein the program, when running, executes the ray-based method for detecting defects in concrete poles as described above.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The defect identification method for concrete poles based on radiographic testing technology enables portable detection of internal hidden dangers in the ground part of concrete poles, which greatly improves operation and maintenance efficiency, saves manpower and material resources, can detect the fracture of the main reinforcement of concrete poles in time, mark the diameter of the reinforcement, guide operation and maintenance personnel to take maintenance measures in time, avoid the expansion of the fault, reduce maintenance costs, and has significant direct economic benefits.

[0017] 2. The method for identifying defects in concrete poles based on radiographic testing technology can promptly and accurately detect safety hazards such as bent steel bars and hollow concrete in concrete poles, effectively avoiding economic losses and safety accidents caused by pole collapse, ensuring power supply safety and reliability, reducing public opinion, maintaining the company's social image, and having significant indirect economic benefits. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0019] Figure 1 This is a flowchart of a method for identifying defects in concrete poles based on X-ray inspection technology. Figure 2 This is a schematic diagram of the installation of a concrete pole defect identification equipment based on X-ray inspection technology (the related power unit is a fixed installation bracket). Figure 3This is a schematic diagram of the installation of a concrete pole defect identification equipment based on X-ray inspection technology (the relevant power unit is a drone). Figure 4 This is a schematic diagram of a concrete pole defect identification equipment based on X-ray inspection technology. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0022] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0023] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0024] The following is combined with Figures 1-4 The embodiments of the present invention will be further described below.

[0025] In one embodiment of the present invention, a method for detecting defects in concrete utility poles based on radiation is provided, comprising the following steps: S1. Install horizontal multi-angle fixtures (including X-ray source and imaging plate). If using insulated poles, pole-climbing robots, etc., install the fixtures and related power devices onto the concrete pole. If using drones, install the fixtures carrying the X-ray source onto the drone, and install the imaging plate onto the concrete pole using clamps.

[0026] S2. Move the tooling to the position to be tested on the pole using methods such as insulating rods, pole-climbing robots, or drones, keeping the imaging plate and the X-ray source at the same horizontal diagonal.

[0027] S3. Remotely start the X-ray machine, penetrate the area to be inspected with X-rays, and read relevant image data from the imaging plate for defect detection.

[0028] S4. Using insulated poles, pole-climbing robots, drones, or other means, the tooling is moved up and down and rotated horizontally to take multi-angle photos of areas where the concrete pole may have defects.

[0029] S5. The obtained defect images are stitched together, and the aforementioned image enhancement and geometric correction methods are used to process the defect images.

[0030] S6. Using artificial intelligence algorithms such as image fault recognition and anomaly detection, combined with an existing defect database, the reliability of concrete poles is evaluated, and the defect type, location, and severity are identified.

[0031] Specifically, to address the problem of identifying internal steel reinforcement fracture and bending defects in the ground portion of concrete poles, a reliability assessment method for the internal state of concrete poles is established by combining methods such as ray image enhancement and two-dimensional geometric correction with deep learning concrete image recognition algorithms, thereby achieving accurate assessment of the internal state of concrete.

[0032] (1) Image enhancement methods Physically driven hybrid filtering Ray attenuation model: combined with formula A layered filter (such as guided filtering + wavelet thresholding denoising) is designed. This formula describes the attenuation law of ray intensity as it propagates through a medium. It shows that the intensity of a ray changes with the propagation distance and the absorption and scattering characteristics of the medium as it passes through it. : Indicates the spatial position of a ray after it has traveled through a medium. Strength at that location, This refers to the initial intensity of the radiation, that is, the intensity of the radiation before it enters the medium under study. : Known as the linear attenuation coefficient, it is an important physical quantity that describes the ability of a medium to absorb and scatter radiation, and its unit is usually m- (per meter). The value depends on the type and density of the medium, as well as the energy of the radiation. Different substances have different abilities to attenuate radiation.

[0033] The guided filtering mentioned above specifically uses edge guidance to suppress scattering noise while preserving the steel reinforcement structure. Wavelet denoising specifically uses multi-scale decomposition to accurately remove high-frequency noise and enhance local contrast.

[0034] The scattering noise and low contrast issues are eliminated using a ray attenuation model. Scattering noise is diffuse noise generated by incoherent processes such as Compton scattering when rays pass through concrete, resulting in uneven image background.

[0035] Dynamic Adaptive Enhancement Framework: Develop a lightweight deep learning model (less than 5M parameters) to achieve dark-field to bright-field image conversion through pairwise training, enhancing the edges of tiny steel bars (contrast improvement of 40%).

[0036] (2) Two-dimensional geometric correction technology In processing images of the internal reinforcement bars of concrete utility poles at ground level, a two-dimensional geometric correction technique based on a single-point perspective model was employed to achieve geometric correction and structural information restoration. This technique is suitable for geometric distortion problems caused by non-standard viewing angle imaging. The processing can be divided into the following key steps: First, an original image of the concrete pole's ground portion is acquired using X-ray or fluoroscopic imaging equipment. This image exhibits significant perspective distortion due to non-standard imaging angles. To correct this distortion, feature information is extracted from the image, primarily including high-contrast areas such as rebar outlines, edges, and intersections. Subsequently, the initial estimated coordinates of these feature points in three-dimensional space are obtained using a three-point localization method, and the corresponding depth is determined with reference to scene settings or physical dimensions. value.

[0037] Next, geometric transformation calculations are performed based on the single-point perspective projection model. The model expression is: This formula is used to represent points in three-dimensional space. , , Mapped to two-dimensional image coordinates , Above, of which: , : These represent the horizontal and vertical coordinates of the reinforcing bars in space; : The distance (depth) from this point to the center of the camera projection; : The focal length of the camera or perspective projection model; , : The coordinates of the corresponding projection point on the image.

[0038] After calculating the ideal projected coordinates of all key points using this formula, pixel mapping is performed in conjunction with the image registration relationship to complete the first round of geometric correction. To further improve the correction accuracy and eliminate nonlinear deformation caused by lens distortion or material inhomogeneity, a thin plate spline (TPS) interpolation algorithm is introduced, which, combined with the relative positional relationship between the X-ray source and the detector, performs deformation compensation for the entire image.

[0039] (3) Construction of the defect map library Multimodal data enhancement: Combining X-ray image signals to generate cross-modal synthetic defect samples. The specific process is as follows: Based on a real ray image background, a simulated physical defect model is injected, and a defect sample conforming to physical laws is synthesized through a ray attenuation model.

[0040] Specific steps for synthesizing defective samples: 1. Acquisition of real radiographic images: Obtain DR images (grayscale images) of defect-free concrete poles as background substrates.

[0041] 2. Physical simulation of defects: Simulation of steel bar fracture, concrete hollowing, and steel bar bending; 3. Defect parametric modeling: Fracture: crack location, width, and cross-sectional angle; Hollowing: cavity volume, depth, and shape; Bending: offset angle, bending radius; 4. Fusion of ray attenuation models; 5. Noise injection.

[0042] Fine-grained labeling system: Defines defect labels (such as rebar breakage, concrete hollowing, and bending angle classification). See the table below for details: (4) Status assessment By utilizing artificial intelligence algorithms such as image fault recognition and anomaly detection, and combining fault and defect risk levels, fault knowledge databases, and historical databases to mine comprehensive internal condition assessment rules for concrete poles, a health index model for concrete poles is established from multiple aspects, thus forming a comprehensive reliability assessment method for concrete poles.

[0043] During the training process of the aforementioned concrete pole health model, the inputs are: the geometric features of the defects (such as the length of the broken steel bars and the area of ​​the hollow concrete); the grayscale features of the defects (such as the average grayscale value of the defect area); and the similarity between the image and the defect template in the image library.

[0044] During model training: Labeled health status data (e.g., normal, slightly damaged, severely damaged) are used to train the model. Model parameters are adjusted so that the model can accurately predict health status based on input features.

[0045] In addition, the step of calculating the health index is included. Specifically, based on the model output, the health index (HI) of the concrete pole is calculated, which can range from 0 to 100. The higher the HI, the better the health condition.

[0046] In one embodiment of the present invention, a X-ray-based defect detection device for concrete utility poles is provided, comprising: Adjustable bracket: The rotating mechanism enables the X-ray inspection device to cover multiple angles (0°~360° horizontal rotation) to meet the inspection needs of poles with different diameters.

[0047] Control: The X-ray machine, imaging plate, and tooling motion axes are linked.

[0048] Modular fixtures: If a climbing device is used to mount the fixture, the X-ray machine and imaging plate are fixed to the multi-angle rotating fixture by mechanical clips. If a drone is used to mount the fixture, the mechanical clips are fixed by manually operating the insulating rod to ensure stable position during the inspection process and avoid image distortion.

[0049] Multi-angle acquisition: By projecting rays at different angles and acquiring images, and combining digital image reconstruction technology to build an internal model, the shadow occlusion problem of single-view imaging is eliminated.

[0050] In summary, the X-ray-based concrete pole defect detection equipment includes a X-ray machine, an imaging plate, a multi-angle horizontal rotating fixture, and a fixed mounting bracket. A climbing device is used to mount the multi-angle detection fixture. The X-ray machine and imaging plate are horizontally adjusted via a multi-angle working mechanism. The fixed mounting bracket is positioned on the concrete pole, or a drone can be used to carry the fixture and X-ray machine. Mechanical clips and the imaging plate are secured manually using an insulating rod, ultimately achieving stable multi-angle X-ray imaging. Standard fixtures enable multi-angle parametric imaging of the ground-level detection points on the concrete pole, solving the problem of inconsistent manual inspections at the same location, achieving unified image capture parameters, and improving detection efficiency.

[0051] In summary, this invention enables portable detection of internal hazards in the ground portion of concrete utility poles, significantly improving maintenance efficiency and saving manpower and material resources. It can promptly detect fractures in the main reinforcing bars of concrete poles, marking the bar diameter to guide maintenance personnel in taking early maintenance measures, preventing the fault from escalating, and reducing repair costs, resulting in significant direct economic benefits. Furthermore, this invention can promptly and accurately detect safety hazards such as bent reinforcing bars and hollow concrete in concrete poles, effectively preventing economic losses and safety accidents caused by pole collapse, ensuring power supply safety and reliability, reducing public opinion negatively, and maintaining the company's social image, resulting in significant indirect economic benefits.

[0052] Those skilled in the art will recognize that the units of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of the invention.

[0053] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical functional division. In actual implementation, there may be other division methods, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored.

[0054] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0055] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0056] 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 method for detecting defects in concrete utility poles based on X-rays, characterized in that, Includes the following steps: Install a horizontal multi-angle fixture that includes at least a radiation source and an imaging plate, and install the fixture and related power devices on the fixture onto a concrete pole, keeping the imaging plate and the radiation source at the same horizontal diagonal. The X-ray machine is remotely started to penetrate the area to be inspected and the X-ray image data is read from the imaging plate. By operating the tooling with a power device, the tooling can be moved up and down and rotated horizontally to take multiple X-ray images of potentially defective parts of the concrete pole. Dark-field and bright-field ray images of the same concrete pole area were collected, paired to form the first dataset, and the ray images of the first dataset were used as the background and injected into the simulated physical defect model to synthesize defect samples. A fault identification model is trained using defect samples, and the trained fault identification model is used to identify whether there are defects in the ray images of concrete poles.

2. The method for detecting defects in concrete utility poles based on radiation according to claim 1, characterized in that, The specific steps for synthesizing defective samples are as follows: Real-world X-ray image acquisition: Obtain DR images of defect-free concrete poles as background substrates; Defect physics simulation: at least simulate steel bar fracture, concrete hollowing, and steel bar bending; Ray attenuation model fusion: The DR image of a defect-free concrete pole is used as the background substrate, and the simulated defects are superimposed on the background and the ground. The fused ray image is then labeled with the corresponding fault tags.

3. The method for detecting defects in concrete utility poles based on radiation according to claim 2, characterized in that, Defect physical simulation specifically involves performing parameterized modeling of defects. Parameterized modeling of defects specifically involves simulating various parameterized defects, with parameters including at least: fracture: crack location, width, cross-sectional angle, voids, cavity volume, depth, and shape.

4. The method for detecting defects in concrete utility poles based on radiation according to claim 2, characterized in that, The process of synthesizing defect samples also includes noise injection, which specifically involves adding noise to the fused ray image after performing ray attenuation model fusion and labeling it with faults.

5. A method for detecting defects in concrete utility poles based on radiation according to any one of claims 2-4, characterized in that, Fault labels should include at least the fault type, fault depth, width, and depth percentage.

6. The method for detecting defects in concrete utility poles based on radiation according to claim 1, characterized in that, After obtaining multiple ray image data, the feature information in the images is first extracted, including the outline of the rebar, edges, and intersections. Then, the initial estimated coordinates of these feature points in three-dimensional space are obtained using the three-point localization method, and the corresponding depth is obtained with reference to the scene settings or physical dimensions. value; Next, geometric transformation calculations are performed based on the single-point perspective projection model. The model expression is as follows: in: , These represent the horizontal and vertical coordinates of the reinforcing bars in space. This is the distance from the point to the center of the camera projection; The focal length of the camera or perspective projection model; , These are the coordinates of the corresponding projection point on the image; After calculating the ideal projected coordinates of all key points, pixel mapping is performed in conjunction with the image registration relationship to complete the first round of geometric correction.

7. The method for detecting defects in concrete utility poles based on radiation according to claim 1, characterized in that, During the construction and training of the fault identification model, at least the geometric features of the defect, the grayscale features of the defect, and the similarity between the image and the defect template in the image database are extracted. Among them, the geometric features of the defect include at least the length of the broken steel bar and the area of ​​the hollow concrete, and the grayscale features of the defect include at least the average grayscale value of the defect area.

8. A X-ray-based defect detection device for concrete utility poles, characterized in that, The method applied to any one of claims 1 to 7 includes: The horizontal multi-angle fixture includes a radiation source and an imaging plate. The radiation source emits radiation that penetrates the concrete pole and is often accompanied by a device to receive the radiation image after penetration. The radiation source and imaging plate are positioned on the same horizontal line and are located at opposite ends of the concrete pole. The relevant power unit is connected to the horizontal multi-angle tooling and is used to carry the tooling to complete the X-ray image acquisition operation. The relevant power unit includes at least one of a fixed mounting bracket and a drone. The fixed mounting bracket is sleeved on the concrete pole and can rotate around the concrete pole. The X-ray source and the imaging plate are respectively connected to the fixed mounting bracket. If the relevant power unit is a drone, the drone is at least connected to the X-ray source, and the imaging plate is driven to rise and fall to the same horizontal line as the X-ray source by connecting to either the telescopic pole or the drone.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the X-ray-based concrete pole defect detection method according to any one of claims 1 to 7.

10. A processor, characterized in that, The processor is used to run a program, wherein the program executes the X-ray-based concrete pole defect detection method according to any one of claims 1 to 7.

Images