Intelligent concrete defect identification method based on potential imaging

Abstract

The invention relates to an intelligent concrete defect identification method based on potential imaging, and the method comprises the following steps: the resistivity of a conductive concrete test piece is tested based on an intelligent detection terminal system to obtain resistivity data, and the intelligent detection terminal system comprises a plurality of electrode pairs arranged in the concrete test piece; performing preliminary geometric positioning on the defect position of the concrete test piece based on the resistivity data to obtain an initial defect position; establishing a corresponding spatial probability distribution model for each pair of electrodes, and calculating the shape factor of the model according to the relative position between the electrode pair and the initial defect position; based on the shape factor, a defect probability detection reconstruction algorithm is adopted to generate a defect feature enhanced fusion image; and performing defect identification on the fused image to obtain a defect identification result. Compared with the prior art, the invention has the advantages of high-precision and high-definition defect imaging, effective enhancement of weak signals, spatialization measurement and the like.

Description

Technical Field

[0001] This invention relates to the field of concrete defect identification, and in particular to an intelligent concrete defect identification method based on potential imaging. Background Technology

[0002] During the long-term service of concrete structures, under the combined effects of loads and environmental factors, defects such as microcracks and voids inevitably develop within them. The initiation and expansion of these defects will seriously threaten the safety and durability of the structure. Therefore, developing efficient and accurate non-destructive testing technologies is crucial for achieving structural health monitoring and early warning.

[0003] Traditional destructive testing methods (such as core sampling and load testing) can directly characterize material properties, but their destructive nature makes them unsuitable for long-term, continuous structural monitoring. Current mainstream nondestructive testing technologies, such as ultrasonic testing, infrared thermography, and spring testing, can identify defects in specific scenarios, but their accuracy is generally susceptible to fluctuations in environmental parameters such as temperature and humidity. Furthermore, they suffer from high dependence on operator experience and limited spatial resolution.

[0004] In recent years, resistivity analysis-based detection methods have attracted much attention due to their non-contact and high-efficiency characteristics. The basic principle is that internal defects in concrete (such as pores and cracks) hinder current conduction, resulting in a significantly higher resistivity in these areas compared to the intact matrix. By measuring the spatial distribution of resistivity, internal defects can be identified and located. However, ordinary concrete has poor conductivity, making it difficult to obtain high-quality signals. Conductive concrete prepared by incorporating functional additives exhibits significantly better conductivity than ordinary concrete, making it an ideal sensing medium with self-sensing capabilities and providing an effective medium for resistivity detection methods.

[0005] Although resistivity measurement methods are theoretically mature, they still face several technical limitations in engineering practice, especially when used for precise imaging with embedded electrode arrays. First, these methods are susceptible to electrical signal noise under complex operating conditions during image reconstruction, and inherent background noise arises during multipath data fusion, leading to blurred images and indistinct defect outlines. Second, balancing electrode quantity and efficiency is difficult; traditional methods require dense electrode placement to obtain high-resolution images, resulting in large data acquisition volumes and low efficiency. Furthermore, currently used detection techniques exhibit weak resistivity responses to small-sized defects, making effective extraction and identification from background noise challenging. These factors limit the practical engineering application of resistivity measurement methods.

[0006] Existing technical solutions: (1) Invention Patent: "A Multi-Dimensional Concrete Cutoff Wall Full-Section Defect Detection Method Based on Electrical Method" (Application No.: CN 119165538 A).

[0007] This invention discloses a multi-dimensional, full-section defect detection method for concrete cut-off walls based on electrical resistivity tomography. It involves pre-embedding five rows of longitudinal electrode probes and multiple layers of transverse electrode probes during the pouring of the cut-off wall to form a detection network. Terminal equipment is used to control the embedded electrode probes and perform defect detection as needed, determining the operational stability of the electrode probes and control terminal. Upon receiving a joint detection request, high-density resistivity method is used from bottom to top for layered detection using transverse electrodes, identifying the first abnormally low-resistivity zone in the entire section. Similarly, electrofluorometry is used from left to right for segmented detection using longitudinal electrodes, identifying the second abnormally low-resistivity zone in the entire section. The first and second abnormally low-resistivity zones are superimposed, and their overlapping area is used as the final determined defect location. The advantages of this method for detecting defects in concrete cut-off walls include: achieving full-section defect identification and location, improving detection accuracy and reliability, reducing damage to the wall structure, and enabling deep defect detection. This proposed scheme, based on a multi-dimensional detection method using electrical resistivity tomography (EDT), can be used for defect detection in anti-seepage walls in hydraulic engineering. It can accurately determine the spatial distribution and range of defects within the wall, providing a basis for subsequent repairs. However, the detection method still has room for improvement in addressing imaging ambiguity and further enhancing accuracy. Secondly, at the data processing level, the current process requires independent inversion calculations and image fusion of horizontal and vertical data; these steps need optimization to improve efficiency.

[0008] (2) Invention patent: "A device and method for testing different cracking conditions inside cement-based materials" (application number: CN 102147387A).

[0009] This invention discloses an apparatus and method for testing different internal cracking conditions in cement-based materials. A standard 40×40×160mm cement-based material specimen is sealed with plastic film and fixed in a flexural testing machine. Two stainless steel electrodes are fixed on two opposite 40×40mm parallel surfaces of the specimen. The electrodes are connected to a potentiostat via working and reference electrodes. During testing, different levels of bending loads are applied to the specimen using the flexural testing machine to induce different cracking conditions. The potentiostat then performs an AC impedance spectroscopy test, and the impedance parameter R collected by a computer reflects the different internal cracking conditions of the cement-based material. This method offers advantages for crack detection in cement-based materials, including: quickly and accurately reflecting different internal cracking conditions, enabling dynamic monitoring of internal cracks, and being simple and practical. This AC impedance spectroscopy-based detection method can be used for the identification and assessment of internal cracks in building materials, determining the stage of crack initiation and propagation. However, data needs to be collected and analyzed intermittently under different load levels; real-time data processing and feature extraction are not yet achieved, and computational efficiency needs improvement. Furthermore, since it does not involve array design and targeted algorithm optimization, its ability to capture and enhance weak signals is limited, which to some extent affects its ability to identify small-sized defects and the accuracy of its location and shape determination.

[0010] In summary, existing technologies suffer from insufficient imaging accuracy and clarity. Furthermore, they are ineffective at enhancing weak signals. Summary of the Invention

[0011] The purpose of this invention is to provide an intelligent concrete defect identification method based on potential imaging to achieve high-precision, high-definition defect imaging and effective enhancement and spatial measurement of weak signals.

[0012] The objective of this invention can be achieved through the following technical solutions: A method for identifying defects in intelligent concrete based on potential imaging, comprising the following steps: The resistivity of conductive concrete specimens is tested using an intelligent detection terminal system to obtain resistivity data. The intelligent detection terminal system includes multiple electrode pairs disposed in the concrete specimens. Preliminary geometric location of defects in concrete specimens is obtained based on resistivity data. A corresponding spatial probability distribution model is established for each pair of electrodes. The positions of the two electrodes are set as the two reference points of the model. The shape factor of the model is calculated based on the relative position between the electrode pair and the initial defect position. A fused image with enhanced defect features is generated based on a defect probability detection and reconstruction algorithm using shape factor. Defect identification is performed on the fused image to obtain the defect identification results.

[0013] Furthermore, the shape factor is: in, β ij It is the shape factor of the model of electrode pair ij; R ij Let be the maximum distance between electrode pair ij and the defect region, where i represents the i-th electrode and j represents the j-th electrode.

[0014] Furthermore, the specific steps for generating a fused image with enhanced defect features based on a defect probability detection and reconstruction algorithm using shape factors are as follows: Calculate the damage index for each electrode pair; Calculate the path distance feature and obtain the constrained path distance feature based on the path distance feature; Defect probability scores are calculated based on damage index and shape factor. The defect probability score and the constrained path distance feature form a spatial probability distribution map. Each spatial probability distribution map is multiplied by its corresponding damage index value to obtain a weighted probability map. All weighted probability maps are linearly superimposed to generate a fused image with enhanced defect features.

[0015] Furthermore, the damage index is: in, t 0 Indicates the transmission time of electrical signals. ΔT The time difference between receiving and transmitting electrical signals. x ij (t) Represents a measurement signal containing defects. y ij (t) Indicates a defect-free reference signal. μ This represents the average value of the corresponding signal.

[0016] Furthermore, the defect probability score is: in, This represents the path distance feature after constraints.

[0017] Furthermore, the constrained path distance feature is: in, Represents the path distance feature. This indicates a uniform parameter.

[0018] Furthermore, the specific steps for defect identification of the fused image to obtain the defect identification results are as follows: Noise images are extracted from fused images enhanced with defect features; Enhanced fused images with enhanced defect features improve the real defect signal; The image after enhancing the real defect signal is compared with the noisy image to obtain the denoised image; Defect identification is performed on the denoised image to obtain the defect identification results.

[0019] Furthermore, the intelligent detection terminal system includes electrodes, wherein multiple sets of transverse electrodes are uniformly arranged along the length of the specimen to form multiple sets of measurement pairs in the x-axis direction; and longitudinal electrodes are symmetrically arranged uniformly in the y-axis direction to form multiple sets of measurement pairs in the y-axis direction.

[0020] Compared with the prior art, the present invention has the following beneficial effects: (1) High-precision, high-definition defect imaging: Existing technologies suffer from blurred images and severe artifacts due to the use of simple abnormal region overlay algorithms and the lack of processing of system imaging noise. This invention introduces a spatial probability distribution model with clear physical meaning to accurately describe the spatial probability distribution of each sensing path, replacing simple geometric overlay, thus providing a more accurate spatial defect localization capability in principle. Furthermore, a defect detection probability reconstruction algorithm is used to fuse the weighted probability map, further enhancing the real defect signal. At the same time, system imaging noise is identified and reduced through image optimization methods.

[0021] (2) Effective enhancement and spatial measurement of weak signals: The rectangular array arrangement adopted in this invention supports resistivity signal acquisition in multiple spatial dimensions, thereby acquiring multi-directional information reflecting the distribution characteristics of defects. At the same time, combined with a dedicated processing algorithm for low signal-to-noise ratio data, the ability to identify weak resistivity changes is significantly improved. This improvement enables the system to not only effectively detect small-sized defects, but also to make more accurate judgments on their spatial location and morphological characteristics, significantly improving the sensitivity and reliability of detection. In addition, this optimized design reduces the number of electrodes required while ensuring high performance, thereby reducing system cost and layout complexity, combining technological advancement with economic efficiency.

[0022] (3) Improved detection efficiency: Existing technologies perform all-electrode pair measurements, generating a large amount of redundant data. This invention innovatively proposes an intersection point localization method based on effective resistivity sensing paths. Through intelligent data analysis, it selects a few effective sensing paths that are most sensitive to and critical to defects from multiple paths, and performs rapid preliminary localization based on their intersection points. This avoids processing all redundant data, allowing computational resources to be concentrated on the most effective information. At the same time, the spatial probability distribution model guides the algorithm to focus on high-probability defect areas, rather than performing full-map calculations. Attached Figure Description

[0023] Figure 1 This is a flowchart of the sensing potential imaging and intelligent assisted concrete defect identification technology of the present invention; Figure 2 This is a schematic diagram of the intelligent detection terminal system of the present invention; Figure 3 The defect identification result based on sensor potential imaging of the present invention is shown below. Figure 3 (a) represents the actual defect area. Figure 3 (b) is the imaging region. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0025] Terminology Explanation: Conductive concrete: A cement-based material with conductive properties prepared by incorporating functional additives. While maintaining structural performance, this material can sense changes in its internal state through variations in its own resistance, making it a key functional material for integrated sensing.

[0026] Functional additives: admixtures used to improve specific properties of concrete. In this invention, they specifically refer to additives that can significantly improve the electrical conductivity of concrete by forming a continuous conductive network in the matrix, thereby enabling the material to have a sensitive response to internal defects.

[0027] Electrode array: This refers to a collection of multiple electrodes pre-embedded within conductive concrete in a specific spatial arrangement. This array serves as the system's signal acquisition interface, and its optimized layout forms a complete measurement network, enabling current to form multiple sensing paths between different electrode pairs. This structural design achieves comprehensive coverage of the measured area, providing a multi-angle, multi-dimensional raw data foundation for subsequent signal processing and imaging analysis.

[0028] Sensing path: A sensing path is a physical concept referring to the actual physical channel through which current flows within concrete from the emitting electrode (excitation end) to the receiving electrode (measurement end) during resistivity measurement. The current distribution in the medium is not an ideal straight line, but rather a three-dimensional spatial field with uneven current density. Each sensing path corresponds to a unique electrode pair, and its resistivity value reflects the overall conductivity of all concrete areas traversed by that path. When defects exist in the path, the resistivity value of that path will change abnormally.

[0029] Spatial Probability Distribution Model: This invention presents a mathematical model for visualizing and quantifying the probability of sensor path detection. The model simulates the sensitive region in space defined by a sensor path of an electrode pair as a specific geometric region with the two electrodes of that pair as reference points. The probability of a defect located at different positions within this region varies. The highest probability occurs near the center line connecting the two electrodes, gradually decreasing with increasing distance from the center line, reaching zero at the region boundary. This model transforms the abstract "line" path into a "surface" region with clear probabilistic meaning, laying the foundation for subsequent image fusion.

[0030] Defect Probability Detection and Reconstruction Algorithm: A dedicated algorithm for defect imaging and localization. It assesses the probability of damage to each path by quantifying and comparing the differences between the signals of each measurement unit in a defective state and a defect-free baseline state. This difference is characterized by a damage index; the closer the value is to the upper limit, the greater the impact of the defect on the path. The algorithm combines the parameter values ​​of each measurement unit with its corresponding spatial distribution model, and through weighted superposition, fuses the damage probability information of all paths into a complete two-dimensional defect probability distribution map, thereby achieving defect imaging enhancement.

[0031] System imaging noise: System imaging noise is a background interference generated by the algorithm after image fusion. It arises because, with a limited number of measurement units, the signal acquisition path coverage density at different spatial locations is inherently uneven. This inherent characteristic of the system layout leads to certain regions in the final fused image background naturally possessing higher baseline values, even if those regions do not contain actual defects. This noise reduces image contrast and interferes with the accurate identification of subtle defects.

[0032] Image optimization method: This invention employs an image post-processing algorithm to suppress imaging noise. Its core is to subtract the inherent background interference image from the original fused image. Specifically, this is achieved by using a specific mathematical processing method to generate a noise image with a pure background, which accurately represents the uneven probability distribution caused by the system layout itself. By performing a difference operation between the original image and this noise image, the background can be effectively balanced, significantly enhancing the contrast of the true defect signal, thereby obtaining a clearer and more reliable final imaging result.

[0033] This invention achieves high-precision, low-noise, and rapid visual detection of internal defects in conductive concrete by using a smaller number of sensors. Specifically, existing non-destructive testing methods for concrete based on resistivity analysis face a core contradiction in engineering applications: improving detection accuracy and resolution typically requires increasing the number of electrodes and collecting a large amount of data, which leads to system complexity, high costs, and low data acquisition and processing efficiency; while reducing the number of electrodes makes it difficult to guarantee the accuracy and imaging quality of defect detection, especially for small-sized defects, and the imaging results are easily affected by noise. Therefore, this invention aims to overcome this technical bottleneck and provide an innovative solution that, without relying on a large number of electrodes, comprehensively improves the positioning accuracy, imaging clarity, and overall efficiency of defect detection through optimized sensor path analysis, model-guided reconstruction algorithms, and image optimization methods.

[0034] Existing non-destructive testing techniques for resistivity mainly suffer from the following problems: The most significant drawback is that imaging accuracy is limited to some extent by artifacts, resulting in insufficient ability to identify and reliably locate small defects. This is because existing technologies employ an "abnormal region overlay" method for defect localization. While this method achieves defect localization by overlaying multipath data, it has limitations in accurately reconstructing the spatial probability distribution of the sensing field. Consequently, the spatial resolution of the reconstructed image is limited, affecting its ability to detect defect boundaries and minute features. Furthermore, the electrode array layout itself introduces specific imaging artifacts. If these systematic errors are not effectively separated, they will significantly reduce the signal-to-noise ratio of the image.

[0035] The data processing workflow needs improvement in computational efficiency. This is because current technologies require high-density, near-complete electrode pair measurements to obtain full-section information, which introduces a significant computational load while generating comprehensive data. This results in increased data acquisition and processing time, impacting overall detection efficiency and making current technologies unsuitable for scenarios requiring rapid response.

[0036] There are still shortcomings in the targeted optimization of array design and signal processing algorithms, which limit its ability to enhance and resolve weak defect signals. This is because a multi-electrode array is not used for spatial signal acquisition, limiting its ability to obtain multi-dimensional spatial distribution information of defects. Simultaneously, the existing technology has limited processing capabilities for low signal-to-noise ratio data, making it difficult to effectively capture the weak resistivity changes caused by small defects. These factors collectively affect the detection rate of small-sized defects and the accuracy of their spatial location and morphology identification in existing technologies.

[0037] (2) In view of the above problems, the purpose of the present invention is as follows: Significantly improves the accuracy and clarity of defect imaging. By introducing a spatial probability distribution model with clear physical meaning to accurately describe the sensing path, and employing a defect probability detection reconstruction algorithm and image optimization methods, the system achieves precise spatial localization and contour restoration of defects, while effectively suppressing system noise, enabling the system to reliably detect sub-centimeter level defects.

[0038] Improve the overall efficiency of the detection process. By using methods such as electrode array design and sensor path analysis, extract the effective sensor path that best reflects changes in resistivity, focus on the most critical data, reduce redundant measurements and calculations, and achieve faster detection speeds while ensuring accuracy.

[0039] This method enhances the sensitivity of detecting internal defects in concrete. By constructing a multi-dimensional electrode array, it effectively captures the subtle resistivity changes caused by defects. Simultaneously, by combining image optimization methods, it significantly improves the signal-to-noise ratio of useful signals to background noise, providing high-quality input data for subsequent accurate identification of small-sized defects and their morphological characteristics.

[0040] The technical solution of this invention is a complete system integrating an intelligent sensing module (electrode array, conductive concrete) and an intelligent recognition and imaging module (path analysis, imaging, noise reduction). Its overall implementation process is as follows: Figure 1 As shown below, each step in the process will be explained in detail.

[0041] (1) Construction of intelligent detection terminal system and acquisition of multi-dimensional resistivity signals: Functional additives are mixed with concrete raw materials according to the designed mix proportions to prepare conductive concrete. The freshly mixed conductive concrete is then poured into a mold. During pouring, several rectangular electrodes are embedded in predetermined positions on the sample, ensuring they are completely embedded in the concrete and maintained at uniform spacing. Figure 2As shown, the electrodes are arranged in a rectangular array along the perimeter of the specimen. Multiple sets of transverse electrodes are evenly arranged along the length of the specimen, forming multiple measurement pairs along the x-axis. Longitudinal electrodes are symmetrically arranged at corresponding positions along the y-axis, together forming a rectangular electrode array suitable for omnidirectional resistivity measurement. This configuration includes multiple measurement points (marked 1, 2, 3, ...), enabling multi-dimensional signal acquisition. After the electrodes are embedded, the conductive concrete specimen undergoes vibration molding and surface curing. Each electrode is connected to an external electrode control unit via a wire. This unit connects two electrodes at a time and transmits the electrical signal to the resistivity measuring device, ultimately forming an intelligent detection terminal with self-sensing capabilities.

[0042] (2) Data analysis and screening based on sensing paths and extraction of defect-sensitive features: Using the intelligent detection terminal constructed in step one, resistivity values ​​of electrodes at different locations are collected pair by pair until the area containing precast defects and the defect-free area are completely covered. Path analysis is performed on the collected resistivity data to identify the current sensing paths corresponding to different electrode pairs. Based on the collected resistivity data, the resistivity change characteristics of each sensing path are analyzed, and the resistivity differences between defective and defect-free paths are compared. Combining the internal conductivity characteristics of concrete, the spatial correlation between the sensing paths and precast defects is derived, sensing paths with abnormal resistivity changes are identified, and path intervals where defects may exist are located. By identifying sensing paths with significantly increased resistivity values, effective resistivity sensing paths that completely cover the defect area are selected. Finally, by calculating the spatial intersections of these effective sensing paths, the preliminary geometric location of the defect is achieved.

[0043] (3) Defect spatial imaging and visualization based on spatial probability distribution model: For each pair of electrodes, a corresponding spatial probability distribution model is established, and the positions of the two electrodes are set as the two reference points of the model. Based on the relative positions of the electrode pair and the defect region, the shape factor of the model is calculated. in, β ij It is the shape factor of the model; R ij This represents the maximum distance (mm) between the electrode pair and the defect area.

[0044] Comparing different β ij Choose the most suitable value. β ij The value ensures that the model coverage is optimally matched with the defect size.

[0045] (4) Image reconstruction and defect feature enhancement using defect probability detection and reconstruction algorithms: 1. Calculate the damage index (SDC) for each electrode pair. By comparing the difference between the defective measurement signal and the defect-free reference signal, quantify the degree of defect impact on each sensing path: in, t 0 Indicates the transmission time of electrical signals. ΔT This is the time difference between receiving and transmitting electrical signals. x ij (t) Represents a measurement signal containing defects. y ij (t) Indicates a defect-free reference signal. μ This represents the average value of the corresponding signal.

[0046] 2. After obtaining the SDC values ​​of all electrode pairs, the image is reconstructed using a defect probability detection and reconstruction algorithm: 3. Correct the initial spatial distribution function and unify the parameters. β Replace with path-related β ij : 4. Use constraint functions to standardize the effective range of the initial spatial distribution, when RD ij ( x , y (Path distance feature) less than or equal to β When +1, take directly. RD ij ( x , y As R ij ( x , y If the threshold is exceeded, then it will be forcibly set to [value]. β +1: 5. Use RD ij ( x , y ) Quantization model reference point (transmitter and receiver electrode positions) x i , y i ), ( x j , y j )) and imaging points ( x ,y Spatial association: 6. Multiply each spatial probability distribution map by its corresponding damage index value to complete the weighted processing of the probability maps.

[0047] 7. Linearly superimpose all weighted probability maps to generate a fused image with enhanced defect features.

[0048] (5) Image optimization and precise defect localization, identification and output of imaging results: 1. For the fused image with enhanced defect features generated in step four, the inherent noise image of the system is extracted by setting an average value Aavg that is independent of the change in the sensing path.

[0049] 2. Image optimization methods are used to optimize imaging data with different weights to enhance the real defect signal: in, P ( x , y ) represents a point ( x , y The expected damage probability at point ) N This represents the total number of sensor electrode pairs. Each sensor electrode pair... ij It has a sensing path, that is P ij ( x , y ).

[0050] 3. By performing a differential operation between the enhanced image of the real defect signal and the system noise image, background noise interference can be effectively suppressed.

[0051] 4. Apply preset defect judgment rules to the clear image after noise reduction. Based on the resistivity anomaly threshold and defect morphology characteristics, identify and output information such as the precise location and size of the defect to complete the entire intelligent detection process.

[0052] The key point of this invention is: (1) Optimized embedded electrode array arrangement scheme: A specific rectangular array arrangement method was designed, including the coordination of horizontal and vertical electrodes, electrode spacing setting and edge positioning.

[0053] (2) Method for preliminary geometric location of defects based on the intersection of the maximum resistivity sensing path: By analyzing resistivity data, effective sensing paths that completely cover the defects are selected, and their spatial intersections are used to achieve rapid and preliminary location of the defects.

[0054] (3) Adaptive spatial probability distribution model: Different spatial probability distribution models are established for each electrode pair, and the model shape is set to dynamically adjust the shape of specific geometric regions in order to accurately describe the spatial probability distribution of the sensing path.

[0055] The advantages of this invention are as follows: (1) High-precision, high-definition defect imaging: Existing technologies suffer from blurred images and severe artifacts due to the use of simple abnormal region overlay algorithms and the lack of processing of system imaging noise. This invention introduces a spatial probability distribution model with clear physical meaning to accurately describe the spatial probability distribution of each sensing path, replacing simple geometric overlay, thus providing a more accurate spatial defect localization capability in principle. Furthermore, a defect detection probability reconstruction algorithm is used to fuse the weighted probability map, further enhancing the real defect signal. At the same time, system imaging noise is identified and reduced through image optimization methods.

[0056] (2) Improved detection efficiency: Existing technologies perform all-electrode pair measurements, generating a large amount of redundant data. This invention innovatively proposes an intersection point localization method based on effective resistivity sensing paths. Through intelligent data analysis, it selects a few effective sensing paths that are most sensitive to and critical to defects from multiple paths, and performs rapid preliminary localization based on their intersection points. This avoids processing all redundant data, allowing computational resources to be concentrated on the most effective information. At the same time, the spatial probability distribution model guides the algorithm to focus on high-probability defect areas, rather than performing full-map calculations.

[0057] (3) Effective enhancement and spatial measurement of weak signals: The rectangular array arrangement adopted in this invention supports resistivity signal acquisition in multiple spatial dimensions, thereby acquiring multi-directional information reflecting the distribution characteristics of defects. At the same time, combined with a dedicated processing algorithm for low signal-to-noise ratio data, the ability to identify weak resistivity changes is significantly improved. This improvement enables the system to not only effectively detect small-sized defects, but also to make more accurate judgments on their spatial location and morphological characteristics, significantly improving the sensitivity and reliability of detection. In addition, this optimized design reduces the number of electrodes required while ensuring high performance, thereby reducing system cost and layout complexity, combining technological advancement with economic efficiency.

[0058] The embedded electrode of this invention can be replaced by a "surface-mounted electrode" or a "drilled-in-place post-installed electrode," suitable for surface inspection of existing structures or structures where damage is not permitted. Compared to this solution: surface electrodes typically have lower measurement depth and signal quality than embedded electrodes and are more susceptible to surface conditions and environmental factors; post-installed electrodes can cause localized damage to the structure, and the contact performance at the electrode-concrete interface may be inferior to electrodes pre-embedded during pouring.

[0059] The functional additives of this invention can be partially or completely replaced with other conductive materials, such as carbon black, graphene, or nickel powder, to meet different cost and performance requirements. Compared to this solution: carbon black is cheaper, but the dosage required to form an effective conductive network is usually higher; graphene has excellent electrical properties, but is expensive and has a complex dispersion process; metal powder has excellent conductivity, but may have problems such as corrosion or sedimentation due to excessive density.

[0060] The following test results are used to demonstrate the beneficial effects and inventiveness of the present invention, in order to quantify the advantages of the present invention: (1) Detection accuracy: Under laboratory conditions, the detection accuracy of prefabricated defects with known location and size is up to 96%.

[0061] (2) Detection sensitivity: The smallest defect diameter that can be reliably identified is up to 6 mm.

[0062] (3) Efficiency and Cost: Compared to traditional all-electrode pair measurements required to achieve the same resolution, this invention reduces the effective data processing volume (and corresponding number of electrodes) by approximately 40%-50% through sensor path filtering, thereby improving computational efficiency and reducing system hardware costs. See the effect diagram below. Figure 3 As shown.

[0063] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a 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 this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for intelligent concrete defect identification based on potential imaging, characterized in that, The method includes the following steps: The resistivity of conductive concrete specimens is tested using an intelligent detection terminal system to obtain resistivity data. The intelligent detection terminal system includes multiple electrode pairs disposed in the concrete specimens. Preliminary geometric location of defects in concrete specimens is obtained based on resistivity data. A corresponding spatial probability distribution model is established for each pair of electrodes. The positions of the two electrodes are set as the two reference points of the model. The shape factor of the model is calculated based on the relative position between the electrode pair and the initial defect position. A fused image with enhanced defect features is generated based on a defect probability detection and reconstruction algorithm using shape factor. Defect identification is performed on the fused image to obtain the defect identification results.

2. The intelligent concrete defect identification method based on potential imaging according to claim 1, characterized in that, The shape factor is: in, β ij It is the shape factor of the model of electrode pair ij; R ij Let be the maximum distance between electrode pair ij and the defect region, where i represents the i-th electrode and j represents the j-th electrode.

3. The intelligent concrete defect identification method based on potential imaging according to claim 1, characterized in that, The specific steps for generating a fused image with enhanced defect features based on the defect probability detection and reconstruction algorithm using shape factor are as follows: Calculate the damage index for each electrode pair; Calculate the path distance feature and obtain the constrained path distance feature based on the path distance feature; Defect probability scores are calculated based on damage index and shape factor. The defect probability score and the constrained path distance feature form a spatial probability distribution map. Each spatial probability distribution map is multiplied by its corresponding damage index value to obtain a weighted probability map. All weighted probability maps are linearly superimposed to generate a fused image with enhanced defect features.

4. The intelligent concrete defect identification method based on potential imaging according to claim 3, characterized in that, The damage index is: in, t 0 Indicates the transmission time of electrical signals. ΔT The time difference between receiving and transmitting electrical signals. x ij (t) Represents a measurement signal containing defects. y ij (t) Indicates a defect-free reference signal. μ This represents the average value of the corresponding signal.

5. The intelligent concrete defect identification method based on potential imaging according to claim 4, characterized in that, The defect probability score is: in, This represents the path distance feature after constraints.

6. The intelligent concrete defect identification method based on potential imaging according to claim 5, characterized in that, The constrained path distance feature is: in, Represents the path distance feature. This indicates a uniform parameter.

7. The intelligent concrete defect identification method based on potential imaging according to claim 1, characterized in that, The specific steps for defect identification in fused images to obtain defect identification results are as follows: Noise images are extracted from fused images enhanced with defect features; Enhanced fused images with enhanced defect features improve the real defect signal; The image after enhancing the real defect signal is compared with the noisy image to obtain the denoised image; Defect identification is performed on the denoised image to obtain the defect identification results.

8. The intelligent concrete defect identification method based on potential imaging according to claim 1, characterized in that, The intelligent detection terminal system includes electrodes, wherein multiple sets of transverse electrodes are uniformly arranged along the length of the specimen to form multiple sets of measurement pairs in the x-axis direction; and longitudinal electrodes are symmetrically arranged uniformly in the y-axis direction to form multiple sets of measurement pairs in the y-axis direction.

9. A smart concrete defect identification device based on potential imaging, comprising a memory, a processor, and a program stored in the memory, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-8.

10. A storage medium having a program stored thereon, characterized in that, When the program is executed, it implements the method as described in any one of claims 1-8.

Images