Underground cable three-dimensional imaging method and device based on X-ray data

Abstract

The invention provides an underground cable three-dimensional imaging method based on X-ray data, and the method comprises the steps: obtaining the X-ray data used for detecting an underground cable, and determining the compressed projection data and an initial attenuation coefficient field of the underground cable according to the X-ray data; based on the compressed projection data, the initial attenuation coefficient field and a preset objective function, the deformation field and the attenuation coefficient field of the underground cable are iteratively updated until the attenuation coefficient field of the underground cable converges, a target attenuation coefficient field is obtained, the objective function is used for realizing sparse reconstruction through cable geometric priori constraint, and in each iteration, the deformation field of the underground cable and the attenuation coefficient field of the underground cable are subjected to sparse reconstruction. Performing parameter optimization by using the updated attenuation coefficient field in the current iteration round; and based on the target attenuation coefficient field and the updated deformation field in each iteration round, performing three-dimensional model fusion to obtain a three-dimensional image of the underground cable. Therefore, efficient, accurate and nondestructive detection of the underground cable can be realized.

Description

Technical Field

[0001] This application relates to the field of power system equipment testing technology, and in particular to a method and apparatus for three-dimensional imaging of underground cables based on X-ray data. Background Technology

[0002] Underground cables, as a crucial infrastructure for urban power transmission, are widely laid along city roads and in industrial parks, and their operational status directly impacts the safety and stability of the power system. However, during long-term burial and operation, underground cables are susceptible to external environmental changes, mechanical stress, and aging, leading to potential hazards such as insulation degradation and metal corrosion. If these hazards are not detected and addressed promptly, they can cause power outages and significant economic losses. Therefore, how to efficiently and accurately perform non-destructive testing and imaging of underground cables has always been a key focus in the power operation and maintenance field.

[0003] Currently, traditional testing methods typically require breaking the cable or relying on external test data to infer its internal condition. However, these methods have significant limitations. Not only is the testing process complex and affects the normal operation of the cable, but improper operation can also cause secondary damage to the cable. Especially when imaging metal structures, these newly generated damages are more prominent in the imaging results, masking the true state of the original defects and increasing the risk of misjudgment. Summary of the Invention

[0004] The purpose of this application is to address at least one of the aforementioned technical deficiencies, particularly the technical deficiencies in the prior art regarding how to efficiently and accurately perform non-destructive testing and imaging of underground cables.

[0005] In a first aspect, this application provides a method for three-dimensional imaging of underground cables based on X-ray data, the method comprising: Acquire X-ray data for detecting underground cables, and determine the compression projection data and initial attenuation coefficient field of the underground cables based on the X-ray data; Based on compressed projection data, initial attenuation coefficient field and preset objective function, the deformation field and attenuation coefficient field of underground cable are iteratively updated until the attenuation coefficient field of underground cable converges to obtain the target attenuation coefficient field. The objective function is used to achieve sparse reconstruction through cable geometric prior constraints, and in each iteration, the parameters are optimized using the attenuation coefficient field updated in the current iteration. Based on the target attenuation coefficient field and the deformation field updated in each iteration, a three-dimensional model is fused to obtain a three-dimensional image of the underground cable.

[0006] In one embodiment, the X-ray data includes the X-ray source energy spectrum distribution, a candidate set of detection angles, and an energy spectrum matrix; the step of determining the compressed projection data and initial attenuation coefficient field of the underground cable based on the X-ray data includes: In the candidate set of detection angles, the optimal projection angle that satisfies the constraint of maximizing incoherence is selected, and compressed projection data is obtained based on the optimal projection angle. Based on the compressed projection data and the energy spectrum matrix, the volume fraction of each material in the underground cable is solved, and the initial attenuation coefficient field is calculated based on the volume fraction of each material and its attenuation coefficient at the reference energy in the X-ray source energy spectrum distribution.

[0007] In one embodiment, the step of iteratively updating the deformation field and attenuation field of the underground cable based on compressed projection data, an initial attenuation coefficient field, and a preset objective function until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field includes: The initial attenuation coefficient field is used as the current attenuation coefficient field of the current iteration. Based on the current attenuation coefficient field, the current deformation field is solved by the preset Horn-Schunck optical flow equation, and the current attenuation coefficient field is corrected using the current deformation field. The objective function is used as the current objective function for the current iteration. The current decay coefficient field is updated using the current objective function, and the parameters of the current objective function are optimized based on the updated current decay field. The updated current attenuation coefficient field is used as the current attenuation coefficient field for the next round, and the optimized current objective function is used as the current objective function for the next round. The deformation field and attenuation coefficient field of the underground cable are updated continuously until the attenuation coefficient field of the underground cable converges, and the target attenuation coefficient field is obtained.

[0008] In one embodiment, the expression for the objective function is:

[0009] in, Indicates data fidelity item, Represents sparse constraint terms. Represents the continuity constraint term. Represents the symmetry constraint term. This represents the current attenuation coefficient field. To compress the observation matrix, To compress the projection data, For sparse constraint term parameters, For parameters of the continuity constraint term, For the symmetry constraint term parameters, For axial gradient operators, It is a radially symmetric constraint.

[0010] In one embodiment, the expression for the Horn-Schunck optical flow equation is:

[0011] in, Indicates the current deformation field. This indicates the current updated decay coefficient field. express Partial derivative with respect to time, Indicates the regularization weight. express Spatial gradient, express Spatial gradient.

[0012] In one embodiment, the expression for parameter optimization of the objective function is:

[0013] in, Indicates the first The constraint parameter, in the first... The update value for the next iteration. This indicates the number of dimensions corresponding to the constraint terms. Indicates the expected value. Indicates the current iteration round The attenuation coefficient field, Indicates the first One constraint item, Represents sparse constraint terms. Indicates continuity constraints, This represents a symmetry constraint.

[0014] In one embodiment, the step of fusing three-dimensional models based on the target attenuation coefficient field and the deformation field updated in each iteration to obtain a three-dimensional image of the underground cable includes: The following expression is used to generate a 3D image of the underground cable:

[0015] in, A three-dimensional image representing underground cables. Represents spatial coordinates, Indicates time, Represents the target attenuation coefficient field. This represents the deformation field that has been updated in each iteration. Indicates a time index.

[0016] Secondly, this application provides a three-dimensional imaging device for underground cables based on X-ray data, the device comprising: The X-ray data acquisition module is used to acquire X-ray data for detecting underground cables and to determine the compression projection data and initial attenuation coefficient field of the underground cables based on the X-ray data. The target attenuation coefficient field determination module is used to iteratively update the deformation field and attenuation coefficient field of the underground cable based on compressed projection data, initial attenuation coefficient field and preset objective function until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field. The objective function is used to achieve sparse reconstruction through cable geometric prior constraints, and in each iteration, the parameters are optimized using the attenuation coefficient field updated in the current iteration. The 3D image determination module is used to fuse 3D models based on the target attenuation coefficient field and the deformation field updated in each iteration to obtain a 3D image of the underground cable.

[0017] Thirdly, this application provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of any of the underground cable three-dimensional imaging methods based on X-ray data in the above embodiments.

[0018] Fourthly, this application provides a computer device, including: one or more processors, and a memory; The memory stores computer-readable instructions, which, when executed by one or more processors, perform the steps of any of the three-dimensional imaging methods for underground cables based on X-ray data in the above embodiments.

[0019] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: The X-ray data-based three-dimensional imaging method for underground cables provided in this application enables efficient, accurate, and non-destructive testing of underground cables. By acquiring X-ray data and determining compressed projection data and initial attenuation coefficient fields, it avoids the need for cable breaking or reliance on external tests in traditional methods, reducing interference with the cable's operational status from the outset. An iterative update mechanism is employed to optimize the cable's deformation field and attenuation coefficient field under the combined effect of compressed projection data, initial attenuation coefficient field, and geometric prior constraint objective function. This results in a target attenuation coefficient field that more closely approximates the cable's true internal state, effectively suppressing false damage features caused by external operations or testing. By optimizing the next round of calculations with updated parameters in each iteration, reconstruction accuracy and convergence speed are improved, avoiding artifacts and misjudgment risks present in traditional imaging. Finally, the fusion of the three-dimensional model based on the target attenuation coefficient field and deformation field achieves clear and realistic three-dimensional imaging of the underground cable's internal structure. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A schematic flowchart of the three-dimensional imaging method for underground cables based on X-ray data provided in the embodiments of this application; Figure 2 A schematic diagram of the structure of the underground cable three-dimensional imaging device based on X-ray data provided in the embodiments of this application; Figure 3 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] This application provides a method for three-dimensional imaging of underground cables based on X-ray data. The following embodiments illustrate this method using a computer device as an example. It is understood that the computer device can be any device with data processing capabilities, including but not limited to a single server, server cluster, personal laptop, desktop computer, etc. Figure 1 As shown, the method may include the following steps: S101: Acquire X-ray data for detecting underground cables, and determine the compression projection data and initial attenuation coefficient field of the underground cables based on the X-ray data.

[0024] X-ray data refers to the raw signal data emitted by X-ray detection equipment and transmitted through underground cables, received and recorded by the detector. This data reflects the absorption and attenuation of X-rays by different structures within the cable. Compressed projection data refers to the data obtained from the X-ray data, which is transformed into a lower-dimensional representation more suitable for reconstruction through mathematical transformations or algorithms. This reduces computational load while preserving key imaging information. The initial attenuation coefficient field refers to the distribution of X-ray attenuation capabilities at various locations within the cable, initially estimated based on the compressed projection data.

[0025] Specifically, the computer equipment first controls the X-ray detection device to scan the target area of ​​the underground cable. This scanning process can be carried out by linear movement along the length of the cable or by multi-angle circumferential scanning to obtain X-ray penetration data from different perspectives. The detector transmits the acquired X-ray signals to the computer equipment in real time. The computer equipment samples and digitizes the raw data to form a standardized X-ray dataset, ensuring the consistency and accuracy of subsequent processing.

[0026] Subsequently, the computer equipment performs compressed projection processing on the acquired X-ray data. Specifically, the computer equipment can execute projection transformation algorithms, such as those based on Fourier transform or Radon transform, to reduce the dimensionality of the multi-dimensional original data into compressed projected data. During this process, the computer equipment can optimize the projection angle and compression ratio based on prior knowledge of the cable's structure, thereby effectively filtering out redundant information, retaining key data features related to internal defects and structure, reducing data volume while maintaining image quality.

[0027] After extracting the compressed projection data, the computer equipment further estimates the initial attenuation coefficient field. In this step, the computer equipment can calculate the preliminary attenuation coefficient distribution at various spatial points inside the cable based on known X-ray physical attenuation models, such as the Bell-Rambert law, combined with the collected projection data. This estimation can be solved using a back-projection algorithm or the least squares method, generating a coarse but complete attenuation coefficient field as the initial input for the 3D imaging of the cable's internal structure. To improve the estimation accuracy, the computer equipment can further utilize a cable material property database to match and calibrate the expected attenuation values ​​for different materials, enhancing the approximation of the initial attenuation field to the actual physical structure.

[0028] By acquiring X-ray data and further determining compressed projection data and the initial attenuation coefficient field, computer equipment can quickly obtain the distribution characteristics of the X-ray response of various parts inside the cable without damaging the cable structure, thus providing a high-quality starting point for imaging iteration. Using compressed projection data processing not only effectively reduces the data processing load and improves computational efficiency, but also ensures that the reconstruction quality is not compromised by dimensionality reduction by preserving key imaging features. The calculation of the initial attenuation coefficient field allows the 3D imaging optimization process to be carried out on a physically reasonable basis that closely approximates the real structure, thereby accelerating convergence, reducing the number of iterations, and avoiding imaging artifacts and misjudgments caused by initial estimation errors. Therefore, non-destructive, efficient, and highly accurate underground cable inspection can be achieved, improving the timeliness and accuracy of cable hazard investigation.

[0029] S102: Based on compressed projection data, initial attenuation coefficient field and preset objective function, the deformation field and attenuation coefficient field of underground cable are iteratively updated until the attenuation coefficient field of underground cable converges to obtain the target attenuation coefficient field. The objective function is used to achieve sparse reconstruction through cable geometric prior constraints, and in each iteration, the parameters are optimized using the attenuation coefficient field updated in the current iteration.

[0030] The deformation field describes the spatial displacement and deformation state of the cable's internal or surface structure due to mechanical stress, aging, or external forces, and is used to characterize the cable's structural health. The objective function is a mathematical function used by the computer equipment to guide the optimization direction during the iteration process. This function, combined with prior geometric information of the cable, is designed into a sparse reconstruction model, ensuring that the imaging results both conform to the physical structure and possess high resolution. The target attenuation coefficient field refers to the attenuation coefficient distribution finally converged after multiple rounds of iterative optimization, used to accurately reflect the true internal state of the cable and generate a three-dimensional image.

[0031] Specifically, the computer device first initiates the parameter update process by calling the iterative reconstruction algorithm module based on the acquired compressed projection data and initial attenuation coefficient field. During initialization, this module uses the initial attenuation coefficient field as input for the first iteration and loads a pre-defined objective function. The objective function incorporates prior geometric knowledge of the cable, such as its circular cross-section, clear internal structure layers, and sparse number of defects, to constrain the distribution of results during optimization and avoid reconstruction results that do not match the actual structure.

[0032] For example, in each iteration, the computer device can first calculate the projection error between the current attenuation coefficient field and the compressed projection data, and input this error as a feedback signal into the objective function. By solving for the minimum value of the objective function, the computer device automatically adjusts the spatial point values ​​of the current attenuation coefficient field, gradually reducing the error. During this process, the computer device simultaneously estimates the deformation field of the underground cable, analyzes the spatial gradient change of the attenuation coefficient, and infers the corresponding physical deformation, thereby compensating for the impact of structural deformation on the projection data and improving the physical accuracy of the attenuation coefficient update.

[0033] Subsequently, after each iteration, the computer equipment performs parameter optimization calculations using the updated decay coefficient field, such as gradient descent, conjugate gradient method, or ADMM (alternating direction multiplier method), to further refine the data field distribution, making it as close as possible to the observed data while ensuring sparsity. For example, it judges the change in the current iteration result compared to the previous round. If the change is less than a preset convergence threshold, the decay coefficient field is considered to have converged, and the iteration ends; otherwise, it automatically enters the next iteration.

[0034] Through multiple rounds of iteration and optimization, the computer equipment finally obtains the target attenuation coefficient field. This result has the characteristics of low error, high contrast and clear details, and can accurately depict the structural integrity inside the underground cable.

[0035] By iteratively updating the deformation and attenuation coefficient fields based on compressed projection data, an initial attenuation coefficient field, and an objective function, computer equipment can effectively integrate prior knowledge of cable structures with actual detection data, achieving high-precision, sparse imaging reconstruction. Employing iterative iteration and parameter optimization not only improves the data field's fit to the actual cable condition but also significantly reduces artifacts and misjudgment risks. Simultaneously, it can adaptively compensate for data distortion caused by cable deformation, thus ensuring the geometric and physical consistency of the reconstruction results. This allows the attenuation coefficient field to quickly converge to its true distribution, shortening computation time, improving detection efficiency, and ultimately achieving non-destructive, rapid, and high-resolution imaging of underground cables.

[0036] S103: Based on the target attenuation coefficient field and the deformation field updated in each iteration, the three-dimensional model is fused to obtain a three-dimensional image of the underground cable.

[0037] Among them, 3D model fusion refers to the process by which computer equipment spatially registers and integrates the target attenuation coefficient field with the deformation fields of each cycle, aiming to synthesize a 3D image of the cable that conforms to the real structural characteristics. The 3D image of the underground cable is a 3D visualization model generated from the fusion result, used to present the internal and external structural details of the cable, and to realize morphological detection and defect location.

[0038] Specifically, the computer equipment first retrieves the target attenuation coefficient field data as the basis for the attenuation characteristics inside the underground cable. Simultaneously, the computer equipment reads the updated deformation field data from each iteration, ensuring that the geometric change information from each iteration is incorporated into subsequent fusion calculations.

[0039] For example, to achieve accurate registration, the computer device can perform coordinate transformation operations on the deformation field of each round, mapping the spatial points in the target attenuation coefficient field according to the displacement vector of the deformation field of that round, generating the attenuation data field corresponding to that round. Subsequently, the computer device performs a fusion operation on the attenuation data fields generated from the above multiple rounds. In a preferred embodiment, the computer device uses a deformation field consistency evaluation algorithm to assign weights to the deformation fields of each round, giving higher weights to data from rounds with stable morphology and appropriately reducing the weights to data from rounds with drastic changes, thereby achieving morphological compensation and smooth transition during fusion. The computer device uses weighted averaging, temporal filtering, or energy minimization-based methods to comprehensively process the multi-round mapping data, ensuring that the fusion result has both geometric consistency and detailed coherence.

[0040] After data fusion is completed, the computer equipment activates the 3D image reconstruction module to convert the fused attenuation data field into a 3D image of the underground cable. The computer equipment preferably uses volume rendering technology to represent the internal attenuation characteristics of the cable, while simultaneously using a surface mesh reconstruction algorithm to restore the external geometric contours of the cable. To improve image clarity, the computer equipment can also perform denoising, edge enhancement, and contrast enhancement processing on the 3D image, resulting in a final 3D image of the underground cable with high resolution and rich detail.

[0041] In the above embodiments, efficient, accurate, and non-destructive testing of underground cables can be achieved. By acquiring X-ray data and determining compressed projection data and initial attenuation coefficient fields, the traditional methods of breaking the cable or relying on external tests are avoided, reducing interference with the cable's operating state from the source. An iterative update mechanism is adopted to optimize the cable's deformation field and attenuation coefficient field under the combined effect of compressed projection data, initial attenuation coefficient field, and geometric prior constraint objective function. This makes the final target attenuation coefficient field closer to the actual internal state of the cable, effectively suppressing false damage features caused by external operations or testing. By using updated parameters to optimize the next round of calculation in each iteration, the reconstruction accuracy and convergence speed are improved, avoiding the risks of artifacts and misjudgments present in traditional imaging. Finally, based on the fusion of the three-dimensional model of the target attenuation coefficient field and deformation field, a clear and realistic three-dimensional image of the internal structure of the underground cable is achieved.

[0042] In one embodiment, the X-ray data includes the X-ray source energy spectrum distribution, a candidate set of detection angles, and an energy spectrum matrix; the step of determining the compressed projection data and initial attenuation coefficient field of the underground cable based on the X-ray data includes: In the candidate set of detection angles, the optimal projection angle that satisfies the constraint of maximizing incoherence is selected, and compressed projection data is obtained based on the optimal projection angle. Based on the compressed projection data and the energy spectrum matrix, the volume fraction of each material in the underground cable is solved, and the initial attenuation coefficient field is calculated based on the volume fraction of each material and its attenuation coefficient at the reference energy in the X-ray source energy spectrum distribution.

[0043] The candidate set of detection angles refers to a set of X-ray irradiation angles pre-set by the computing device for projection data acquisition. The incoherence maximization constraint is a mathematical criterion used to optimize the angle selection process, aiming to ensure maximum complementarity of observation information between selected projection angles, reducing redundancy and improving reconstruction accuracy. The optimal projection angle is the projection angle selected from the candidate set that best meets the incoherence maximization constraint, used for actual compressed projection data acquisition. Compressed projection data refers to the X-ray projection signal acquired based on the optimal projection angle and processed through data compression, used to reduce data volume while retaining core imaging information. The energy spectrum matrix is ​​matrix data characterizing the energy distribution of the X-ray source at different energies, a key input for solving material composition. The volume fraction refers to the proportion of different materials in the total volume of the underground cable. The initial attenuation coefficient field refers to the spatial attenuation coefficient distribution of the underground cable calculated based on the material volume fraction and X-ray energy spectrum reference data, used as the starting data field for image reconstruction and optimization iteration.

[0044] Specifically, the computing device first loads a candidate set of detection angles, which includes multiple preset ray illumination directions. Based on the constraint of maximizing incoherence, the computing device calculates the pairwise information overlap of all candidate angle pairs and selects the set of angles with the highest information complementarity and lowest redundancy as the optimal projection angles. In a preferred embodiment, the computing device uses a coherence metric based on determinant or an optimization function based on subspace orthogonality to perform a combined screening of the candidate angle set, ensuring that the finally selected optimal projection angle set meets the requirements of wide information distribution and sufficient reconstruction conditions.

[0045] Subsequently, the computer equipment controls the X-ray acquisition unit to perform multi-angle scanning of the underground cable target area according to the aforementioned optimal projection angle, acquiring raw projection data. To reduce data transmission and computational load, the computer equipment performs compression processing on the acquired raw projection data, such as using sparse coding, subsampling, or transform domain compression methods, to generate compressed projection data.

[0046] Next, the computer device loads pre-stored energy spectrum matrix data, representing the energy distribution characteristics of the X-ray source in each energy band. Based on the compressed projection data and the energy spectrum matrix, the computer device performs material decomposition calculations to solve for the volume fraction of each component material in the underground cable. In a preferred embodiment, the device employs a least-squares fitting or iterative reprojection algorithm, utilizing the linear or nonlinear relationship between the energy spectrum matrix and the projection data to solve for the volume distribution of each material.

[0047] After the volume fraction is calculated, the computer equipment performs a weighted calculation based on the attenuation coefficient data of each material at the X-ray source reference energy, according to the volume fraction of each material, to generate an initial attenuation coefficient field. The spatial distribution of this field data is aligned with the cable structure, reflecting the overall attenuation characteristics of the composite material at each spatial location point.

[0048] In one example, firstly, taking advantage of the sparse and geometrically symmetrical internal structure of the cable, multi-angle compressed sensing data acquisition is implemented, with the energy spectrum distribution of an X-ray source as the input. And the candidate set of detection angles Θ. Traditional CT scans require the Shannon-Nyquist sampling theorem, typically necessitating a large number of projection angles. However, considering the sparse and regular geometric features of the internal structure of the cable, this example, based on compressed sensing theory, significantly reduces the number of projection angles to only 30% of the conventional acquisition requirements, thereby optimizing data acquisition. In mathematical modeling, the attenuation coefficient distribution function of the target cable is denoted as... It describes the attenuation characteristics of the cable's internal space in cylindrical coordinates. Through coordinate transformation, it can be... Convert to Cartesian coordinate system This facilitates integration with projection models.

[0049] The projection data model on which the data acquisition is based is:

[0050] in, Indicates the projection angle. These are the position parameters on the detector. For the integral path along this projection angle and position, To conform to a normal distribution The noise term reflects the actual detection error.

[0051] This example constructs a compressed observation matrix that satisfies the Finite Isometric Property (RIP) by optimizing the projection angle selection strategy. ,in much smaller ,accomplish:

[0052] To further improve the reconstruction effect, the angle selection is optimized so that Φ and the cable sparse basis Ψ (such as wavelet basis) satisfy the incoherence maximization condition, and the formula is:

[0053] in, sparse base The first in basis vectors To compress the observation matrix The first in row vectors. This represents the total number of pixels corresponding to the spatial resolution. Through optimization... To achieve the optimal compressed observation matrix and output compressed projection data:

[0054] in, The cable attenuation coefficient field in three-dimensional space. This is the noise term.

[0055] After compressed sensing data acquisition is completed, multi-spectral material decomposition is performed, with the input data being the compressed projection data obtained above. and energy spectrum matrix Based on the attenuation coefficients of multi-energy X-rays on different materials With energy The changing characteristics allow for the separation of the internal material components of the cable. Cables are typically composed of a metallic conductor (such as copper, Cu) and an insulation layer (such as cross-linked polyethylene, XLPE), and their attenuation characteristics differ significantly. Let the first... The volume fraction of the material is Then the overall attenuation coefficient satisfies:

[0056] in, For the first This material in energy The attenuation coefficient below, This represents the number of material types. Energy spectral matrix. The construction is as follows: Indicates the first The material in the first The attenuation coefficient at each energy spectrum sampling point is obtained by solving the volume fraction vector. To achieve material decomposition, the specific solution model is as follows:

[0057] in, This is a regularization parameter used to enhance sparsity and suppress the noise sensitivity of the solution. To improve the decomposition accuracy, the K-edge characteristics of metallic materials (such as copper) are also incorporated. For example, the K-edge of Cu is located at 8.98 keV. By selecting an energy spectrum that covers this energy point, the material discrimination capability is enhanced.

[0058] Finally, the volume fraction of each material was calculated. And calculate the initial values ​​of the material distribution:

[0059] in, The selected reference energy is used to unify the attenuation coefficients of various materials, so that the obtained This represents the initial attenuation coefficient field of the cable under the reference energy.

[0060] In this embodiment, by selecting the optimal projection angle that satisfies the constraint of maximizing incoherence from the candidate set of detection angles, the computer equipment can ensure that the acquired data contains information with strong complementarity and low redundancy, thereby effectively improving the accuracy of material decomposition and image reconstruction. Using the energy spectrum matrix to perform multi-energy spectrum decomposition on the projection data enables the computer equipment to accurately solve for the volume fraction of different materials in the underground cable, achieving precise differentiation at the material composition level. The initial attenuation coefficient field calculated based on the volume fraction and reference attenuation coefficient not only provides an accurate starting point for iterative optimization but also improves the convergence speed and stability of the overall reconstruction process. Overall, this method ensures the efficiency and reliability of the data acquisition and reconstruction process, significantly improving the imaging quality and detection accuracy of underground cable detection.

[0061] In one embodiment, the step of iteratively updating the deformation field and attenuation coefficient field of the underground cable based on compressed projection data, an initial attenuation coefficient field, and a preset objective function until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field includes: The initial attenuation coefficient field is used as the current attenuation coefficient field of the current iteration. Based on the current attenuation coefficient field, the current deformation field is solved by the preset Horn-Schunck optical flow equation, and the current attenuation coefficient field is corrected using the current deformation field. The objective function is used as the current objective function for the current iteration. The current decay coefficient field is updated using the current objective function, and the parameters of the current objective function are optimized based on the updated current decay field. The updated current attenuation coefficient field is used as the current attenuation coefficient field for the next round, and the optimized current objective function is used as the current objective function for the next round. The deformation field and attenuation coefficient field of the underground cable are updated continuously until the attenuation coefficient field of the underground cable converges, and the target attenuation coefficient field is obtained.

[0062] The Horn-Schunck optical flow equation is a mathematical model used in computer vision to estimate the displacement vector field between images or volume data.

[0063] Specifically, the computer device first loads the initial attenuation coefficient field obtained through multi-energy spectral material decomposition as the initial value for the current iteration. Then, the computer device caches this attenuation coefficient field in a high-performance storage unit and calls the deformation solving module. Based on the spatial gradient characteristics of the current attenuation coefficient field, it uses the preset Horn-Schunck optical flow equation to calculate and solve for the current deformation field. This process can generate a complete three-dimensional deformation field by traversing a three-dimensional voxel mesh and applying the optical flow constraint equation to each voxel point to calculate its displacement vector.

[0064] Next, the computer equipment uses the obtained current deformation field to correct the current attenuation coefficient field. Specifically, the computer equipment performs coordinate mapping and interpolation calculations on the voxel data in the attenuation coefficient field based on the displacement vector of each voxel point in the deformation field, thereby achieving deformation correction. This process can be executed in parallel on the GPU using linear interpolation or B-spline interpolation to ensure processing speed and accuracy.

[0065] Subsequently, the computer device uses the pre-defined objective function as the optimization function for the current iteration and further updates the deformation-corrected attenuation coefficient field. This step adjusts the voxel values ​​in the attenuation coefficient field by minimizing the objective function value, making them better match the projected data and geometric priors. During this process, the computer device dynamically adjusts the weight parameters in the objective function, such as the regularization coefficient and the weight of the data fitting term, to achieve a balance between convergence speed and reconstruction quality.

[0066] After completing the above updates, the computer device sets the updated attenuation coefficient field as the current attenuation coefficient field for the next round, and sets the objective function adjusted according to the previous optimization process as the current objective function for the next round. The computer device repeats this series of actions—deformation field solution, attenuation coefficient field correction, and objective function update—forming an iterative loop. At the end of each iteration, the computer device determines whether to continue iterating based on preset convergence conditions, such as the change in the objective function being lower than a threshold or the number of iterations reaching an upper limit, until the cable's attenuation coefficient field converges, and the final output is the target attenuation coefficient field.

[0067] In this embodiment, alternating the calculation of the deformation field and the updating of the attenuation coefficient field effectively couples the deformation information of the underground cable with its material attenuation characteristics. By using the Horn-Schunck optical flow equation to solve the deformation field, the computer equipment can accurately capture the structural displacement of the cable caused by external forces or aging, thereby spatially correcting the attenuation coefficient field and improving the geometric consistency of the reconstruction results. Furthermore, by iteratively updating the corrected attenuation coefficient field based on the objective function, the reconstruction results can continuously approach the true projection data, improving imaging accuracy. In addition, the dynamic optimization of the objective function allows the equipment to adaptively adjust parameters according to the current state of the attenuation coefficient field, avoiding overfitting or underfitting. Therefore, this process achieves synergistic optimization of the deformation field and the attenuation coefficient field, ultimately obtaining a more accurate target attenuation coefficient field with a more realistic material distribution, improving the reliability and accuracy of underground cable detection.

[0068] In one embodiment, the expression for the objective function is:

[0069] in, Indicates data fidelity item, Represents sparse constraint terms. Represents the continuity constraint term. Represents the symmetry constraint term. This represents the current attenuation coefficient field. To compress the observation matrix, To compress the projection data, For sparse constraint term parameters, For parameters of the continuity constraint term, For the symmetry constraint term parameters, For axial gradient operators, It is a radially symmetric constraint.

[0070] The expression for the radial symmetry constraint is:

[0071] Specifically, the objective function is designed to guide the reconstruction and optimization of the attenuation coefficient field of underground cables, with each term having a clear functional division. The data fidelity term measures the distance between the current attenuation coefficient field projected by the projection operator and the actual projected measurement data. Minimizing this term ensures consistency between the reconstructed results and the actual observation data, thereby improving reconstruction accuracy. The sparsity constraint term introduces a norm to promote the sparsity of the attenuation coefficient field in certain transform domains (such as wavelet or gradient domains), helping to suppress noise and remove artifacts while improving the clarity of cable structural features. This term is suitable for data reconstruction in complex environments and helps improve the system's robustness to low signal-to-noise ratio data. The continuity constraint term constrains the smoothness of the attenuation coefficient field in the axial direction, suppressing local abrupt changes and excessive oscillations, thereby improving the continuity and stability of the reconstruction model in the longitudinal dimension and adapting to the morphological extension characteristics of underground cables. The symmetry constraint term minimizes the distance between the current field and the deformation field correction result. This term ensures that the attenuation coefficient field and deformation field information remain coupled and consistent, helping to improve the reconstruction model's ability to capture cable deformation features, thereby improving the accuracy of deformation detection.

[0072] In summary, the objective function achieves synergistic optimization across multiple dimensions, including data fitting, noise suppression, spatial smoothing, and deformation consistency. On one hand, it ensures high consistency between the reconstructed results and the actual measurement data; on the other hand, it effectively improves reconstruction quality and suppresses noise and artifacts through a regularization strategy. Furthermore, the deformation coupling mechanism enhances the system's responsiveness to changes in the underground cable structure, thereby significantly improving the imaging accuracy, structural reconstruction capability, and application reliability of the underground cable detection system.

[0073] In one embodiment, the expression for the Horn-Schunck optical flow equation is:

[0074] in, Indicates the current deformation field. This indicates the current updated decay coefficient field. express Partial derivative with respect to time, Indicates the regularization weight. express Spatial gradient, express Spatial gradient.

[0075] Specifically, the Horn-Schunck optical flow equation is used to estimate the deformation field of the target, i.e. the attenuation coefficient field of the underground cable, in the image sequence between consecutive frames, and has the ability to accurately describe continuous displacement changes.

[0076] This is the optical flow constraint term, also known as the brightness consistency term. This term is based on the optical flow assumption that the brightness value of the target point remains constant over time during deformation. Minimizing this term effectively ensures that the solved deformation field accurately reflects the variation of the underground cable attenuation coefficient field over time or between frames, thereby improving the accuracy of deformation estimation. This is a smoothness regularization term for the optical flow field. By constraining the continuity of the deformation field in the spatial dimension, it suppresses abrupt changes caused by local noise or measurement errors during the solution process, ensuring the overall smoothness and physical rationality of the deformation field results. The weighting parameters are... This is used to balance the weight relationship between data fitting terms and smoothness constraint terms, in order to adapt to deformation estimation needs in different scenarios.

[0077] In summary, the Horn-Schunck optical flow equation improves the consistency between the estimated deformation field and the actual attenuation coefficient field through brightness consistency constraints, helping to accurately characterize the local displacement and morphological changes of underground cables. Furthermore, by introducing smoothness constraints, it effectively suppresses the impact of noise interference on the deformation field calculation, thereby improving the spatial continuity and robustness of the estimation results. Overall, this method achieves a balance between high accuracy and high robustness in deformation estimation, providing a reliable foundation for the correction and reconstruction of the attenuation coefficient field of underground cables.

[0078] In one embodiment, the expression for parameter optimization of the objective function is:

[0079] in, Indicates the first The constraint parameter, in the first... The update value for the next iteration. This indicates the number of dimensions corresponding to the constraint terms. Indicates the expected value. Indicates the current iteration round The attenuation coefficient field, Indicates the first One constraint item, Represents sparse constraint terms. Indicates continuity constraints, This represents a symmetry constraint.

[0080] Specifically, this formula is used to adaptively optimize each regularization parameter in the objective function in order to improve the balance of the objective function among different constraint terms and the overall solution effect.

[0081] Overall, this parameter optimization achieves adaptive dynamic adjustment of the weights of each regularization term, avoiding the risks of underfitting or overfitting caused by traditional fixed parameter settings. On the other hand, it improves the numerical stability and convergence speed in the objective function solution process, which helps to accurately reconstruct the attenuation coefficient field of underground cables while maintaining sparsity, continuity and symmetry characteristics, thereby improving the overall detection and reconstruction accuracy.

[0082] In one embodiment, the step of fusing three-dimensional models based on the target attenuation coefficient field and the deformation field updated in each iteration to obtain a three-dimensional image of the underground cable includes: The following expression is used to generate a 3D image of the underground cable:

[0083] in, A three-dimensional image representing underground cables. Represents spatial coordinates, Indicates time, Represents the target attenuation coefficient field. This represents the deformation field that has been updated in each iteration. Indicates a time index.

[0084] Specifically, this formula enables dynamic sampling of the original attenuation coefficient field based on the cumulative deformation displacement of the cable at different time steps, thereby generating a three-dimensional image with temporal variation characteristics. This generation process effectively integrates spatial attenuation characteristics and temporal deformation information, allowing the output image to accurately reflect the structural deformation of the cable during long-term burial due to ground movement, settlement, or external forces.

[0085] Overall, this 3D image generation, on the one hand, enhances the ability to capture the actual deformation dynamics of the cable by superimposing deformation fields at multiple times, and realizes the visualization of the temporal morphological changes of the cable; on the other hand, the generated 3D image has high spatial continuity and temporal consistency.

[0086] The following describes the underground cable three-dimensional imaging device based on X-ray data provided in the embodiments of this application. The underground cable three-dimensional imaging device based on X-ray data described below can be referred to in correspondence with the underground cable three-dimensional imaging method based on X-ray data described above. Figure 2 As shown, this application provides a three-dimensional imaging device for underground cables based on X-ray data. The device includes: X-ray data acquisition module 201 is used to acquire X-ray data for detecting underground cables and to determine the compression projection data and initial attenuation coefficient field of the underground cables based on the X-ray data. The target attenuation coefficient field determination module 202 is used to iteratively update the deformation field and attenuation coefficient field of the underground cable based on compressed projection data, initial attenuation coefficient field and preset objective function until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field. The objective function is used to achieve sparse reconstruction through cable geometric prior constraints, and in each iteration, the parameters are optimized using the attenuation coefficient field updated in the current iteration. The three-dimensional image determination module 203 is used to perform three-dimensional model fusion based on the target attenuation coefficient field and the deformation field updated in each iteration to obtain a three-dimensional image of the underground cable.

[0087] In one embodiment, the X-ray data includes the X-ray source energy spectrum distribution, a candidate set of detection angles, and an energy spectrum matrix; the X-ray data acquisition module 201 includes: The compressed projection data acquisition unit is used to select the optimal projection angle that satisfies the incoherence maximization constraint from the candidate set of detection angles, and acquire compressed projection data based on the optimal projection angle. The initial attenuation coefficient field calculation unit is used to solve the volume fraction of each material in the underground cable based on the compressed projection data and the energy spectrum matrix, and to calculate the initial attenuation coefficient field based on the volume fraction of each material and its attenuation coefficient at the reference energy in the X-ray source energy spectrum distribution.

[0088] In one embodiment, the target attenuation coefficient field determination module 202 includes: The current attenuation coefficient field correction unit is used to take the initial attenuation coefficient field as the current attenuation coefficient field of the current iteration, solve the current deformation field according to the current attenuation coefficient field through the preset Horn-Schunck optical flow equation, and correct the current attenuation coefficient field using the current deformation field. The objective function parameter optimization unit is used to take the objective function as the current objective function of the current iteration, update the corrected current decay coefficient field with the current objective function, and optimize the parameters of the current objective function based on the updated current decay field. The target attenuation coefficient field determination unit is used to take the updated current attenuation coefficient field as the current attenuation coefficient field for the next round, and the optimized current objective function as the current objective function for the next round, and continue to update the deformation field and attenuation coefficient field of the underground cable until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field.

[0089] In one embodiment, the expression for the objective function is:

[0090] in, Indicates data fidelity item, Represents sparse constraint terms. Represents the continuity constraint term. Represents the symmetry constraint term. This represents the current attenuation coefficient field. To compress the observation matrix, To compress the projection data, For sparse constraint term parameters, For parameters of the continuity constraint term, For the symmetry constraint term parameters, For axial gradient operators, It is a radially symmetric constraint.

[0091] In one embodiment, the expression for the Horn-Schunck optical flow equation is:

[0092] in, Indicates the current deformation field. This indicates the current updated decay coefficient field. express Partial derivative with respect to time, Indicates the regularization weight. express Spatial gradient, express Spatial gradient.

[0093] In one embodiment, the expression for parameter optimization of the objective function is:

[0094] in, Indicates the first The constraint parameter, in the first... The update value for the next iteration. This indicates the number of dimensions corresponding to the constraint terms. Indicates the expected value. Indicates the current iteration round The attenuation coefficient field, Indicates the first One constraint item, Represents sparse constraint terms. Indicates continuity constraints, This represents a symmetry constraint.

[0095] In one embodiment, the three-dimensional image determination module 203 includes: A 3D image determination unit is used to generate a 3D image of the underground cable using the following expression:

[0096] in, A three-dimensional image representing underground cables. Represents spatial coordinates, Indicates time, Represents the target attenuation coefficient field. This represents the deformation field that has been updated in each iteration. Indicates a time index.

[0097] In one embodiment, this application also provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the underground cable three-dimensional imaging method based on X-ray data as described in any of the above embodiments.

[0098] In one embodiment, this application also provides a computer device storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the three-dimensional imaging method for underground cables based on X-ray data as described in any of the above embodiments.

[0099] Indicatively, such as Figure 3 As shown, Figure 3 This is a schematic diagram of the internal structure of a computer device 300 provided in an embodiment of this application. The computer device 300 can be provided as a server. (Refer to...) Figure 3 The computer device 300 includes a processing component 302, which further includes one or more processors, and memory resources represented by memory 301 for storing instructions, such as application programs, that can be executed by the processing component 302. The application programs stored in memory 301 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 302 is configured to execute instructions to perform the underground cable three-dimensional imaging method based on X-ray data of any of the above embodiments.

[0100] The computer device 300 may also include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate on an operating system stored in memory 301, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or similar.

[0101] Those skilled in the art will understand that Figure 3The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0102] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this document, "a," "an," "the," "the," and "its" may also include plural forms unless the context clearly indicates otherwise. "Multiple" refers to at least two, such as 2, 3, 5, or 8, etc. "And / or" includes any and all combinations of the related listed items.

[0103] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0104] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for three-dimensional imaging of underground cables based on X-ray data, characterized in that, The method includes: Acquire X-ray data for detecting underground cables, and determine the compression projection data and initial attenuation coefficient field of the underground cables based on the X-ray data; Based on the compressed projection data, the initial attenuation coefficient field, and the preset objective function, the deformation field and attenuation coefficient field of the underground cable are iteratively updated until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field. The objective function is used to achieve sparse reconstruction through cable geometric prior constraints, and in each iteration, the parameters are optimized using the attenuation coefficient field updated in the current iteration. Based on the target attenuation coefficient field and the deformation field updated in each iteration, a three-dimensional model is fused to obtain a three-dimensional image of the underground cable.

2. The method for three-dimensional imaging of underground cables based on X-ray data according to claim 1, characterized in that, The X-ray data includes the X-ray source energy spectrum distribution, candidate set of detection angles, and energy spectrum matrix; The step of determining the compressed projection data and initial attenuation coefficient field of the underground cable based on the X-ray data includes: In the candidate set of detection angles, the optimal projection angle that satisfies the constraint of maximizing incoherence is selected, and the compressed projection data is obtained based on the optimal projection angle. Based on the compressed projection data and the energy spectrum matrix, the volume fraction of each material in the underground cable is determined, and the initial attenuation coefficient field is calculated based on the volume fraction of each material and its attenuation coefficient at the reference energy in the X-ray source energy spectrum distribution.

3. The method for three-dimensional imaging of underground cables based on X-ray data according to claim 1, characterized in that, The step of iteratively updating the deformation field and attenuation field of the underground cable based on the compressed projection data, the initial attenuation coefficient field, and the preset objective function until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field includes: The initial attenuation coefficient field is used as the current attenuation coefficient field of the current iteration. Based on the current attenuation coefficient field, the current deformation field is solved by the preset Horn-Schunck optical flow equation, and the current attenuation coefficient field is corrected using the current deformation field. The objective function is used as the current objective function for the current iteration. The current attenuation coefficient field is updated using the current objective function, and the parameters of the current objective function are optimized based on the updated current attenuation field. The updated current attenuation coefficient field is used as the current attenuation coefficient field for the next round, and the optimized current objective function is used as the current objective function for the next round. The deformation field and attenuation coefficient field of the underground cable are updated continuously until the attenuation coefficient field of the underground cable converges, thus obtaining the target attenuation coefficient field.

4. The method for three-dimensional imaging of underground cables based on X-ray data according to claim 3, characterized in that, The expression for the objective function is: in, Indicates data fidelity item, Represents sparse constraint terms. Represents the continuity constraint term. Represents the symmetry constraint term. This represents the current attenuation coefficient field. To compress the observation matrix, For the compressed projection data, For sparse constraint term parameters, For parameters of the continuity constraint term, For the symmetry constraint term parameters, For axial gradient operators, It is a radially symmetric constraint.

5. The method for three-dimensional imaging of underground cables based on X-ray data according to claim 3, characterized in that, The expression for the Horn-Schunck optical flow equation is as follows: in, Indicates the current deformation field. This indicates the current updated decay coefficient field. express Partial derivative with respect to time, Indicates the regularization weight. express Spatial gradient, express Spatial gradient.

6. The method for three-dimensional imaging of underground cables based on X-ray data according to claim 4, characterized in that, The expression for parameter optimization of the objective function is as follows: in, Indicates the first The constraint parameter, in the first... The update value for the next iteration. This indicates the number of dimensions corresponding to the constraint terms. Indicates the expected value. Indicates the current iteration round The attenuation coefficient field, Indicates the first One constraint item, Represents sparse constraint terms. Indicates continuity constraints, This represents a symmetry constraint.

7. The method for three-dimensional imaging of underground cables based on X-ray data according to claim 1, characterized in that, The step of fusing the three-dimensional model based on the target attenuation coefficient field and the deformation field updated in each iteration to obtain a three-dimensional image of the underground cable includes: The following expression is used to generate a three-dimensional image of the underground cable: in, This represents a three-dimensional image of the underground cable. Represents spatial coordinates, Indicates time, Represents the target attenuation coefficient field. This represents the deformation field that has been updated in each iteration. Indicates a time index.

8. A three-dimensional imaging device for underground cables based on X-ray data, characterized in that, The device includes: The X-ray data acquisition module is used to acquire X-ray data for detecting underground cables, and to determine the compression projection data and initial attenuation coefficient field of the underground cable based on the X-ray data. The target attenuation coefficient field determination module is used to iteratively update the deformation field and attenuation coefficient field of the underground cable based on the compressed projection data, the initial attenuation coefficient field and the preset objective function, until the attenuation coefficient field of the underground cable converges to obtain the target attenuation coefficient field. The objective function is used to achieve sparse reconstruction through cable geometric prior constraints, and in each iteration, the parameters are optimized using the attenuation coefficient field updated in the current iteration. The three-dimensional image determination module is used to perform three-dimensional model fusion based on the target attenuation coefficient field and the deformation field updated in each iteration to obtain a three-dimensional image of the underground cable.

9. A storage medium, characterized in that: The storage medium stores computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the three-dimensional imaging method for underground cables based on X-ray data as described in any one of claims 1 to 7.

10. A computer device, characterized in that, include: One or more processors, and memory; The memory stores computer-readable instructions, which, when executed by the one or more processors, perform the steps of the three-dimensional imaging method for underground cables based on X-ray data as described in any one of claims 1 to 7.

Images