Three-dimensional dynamic reconstruction method and system for long-gap discharge channel and storage medium

Abstract

The invention discloses a three-dimensional dynamic reconstruction method for a long-gap discharge channel, and the method comprises the steps: collecting a multi-view synchronous image sequence of a discharge process through at least two cameras which are disposed on the same plane and are synchronously triggered, carrying out the graying, binaryzation and time correction processing of a synchronous image group at each time point, and carrying out the calculation of the multi-view synchronous image sequence. Obtaining a binarized image with time alignment; performing visual angle correction on the image, and extracting discharge channel edge points line by line in the corrected image to form an edge point set; then, constructing an overdetermined equation set by utilizing the edge point set based on epipolar constraint, and solving through singular value decomposition to obtain a three-dimensional space coordinate set; and finally, generating a three-dimensional static model of a single time point by calculating central points and radiuses of sections with different heights, and combining the static models of continuous time points in sequence to finally reconstruct a three-dimensional dynamic model of the discharge channel. According to the method, high-precision and high-efficiency three-dimensional dynamic reduction of the transient evolution process of the discharge channel is realized.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional reconstruction technology of discharge channels, and in particular to a method, system and storage medium for three-dimensional dynamic reconstruction of long-gap discharge channels. Background Technology

[0002] The study of long-gap discharge phenomena is of vital importance in fields such as power system insulation design, equipment safety protection, and the exploration of the physical mechanisms of lightning. To gain a deeper understanding of the discharge development process, researchers typically use high-speed imaging technology to capture this transient phenomenon.

[0003] In existing technologies, a mainstream approach relies on one or more high-speed cameras to acquire two-dimensional image sequences of the discharge process. While this method offers high temporal resolution and can capture instantaneous morphological changes in the discharge channel, its fundamental drawback lies in the lack of spatial information. Two-dimensional images are merely projections of three-dimensional objects onto a plane, failing to accurately depict the complex three-dimensional structures of the discharge channel in real space, such as twists, bifurcations, and rotations. This limits in-depth analysis of the discharge path and energy distribution.

[0004] To overcome the limitations of two-dimensional imaging, some studies have introduced multi-view vision and three-dimensional reconstruction techniques. However, these traditional three-dimensional methods face severe challenges in practical applications. Most of them rely on complex gray-level correlation matching algorithms to find corresponding points in images from different viewpoints. This not only leads to a sharp increase in computational load with image resolution and low processing efficiency, but also makes them highly susceptible to interference from plasma flicker and background noise generated during the discharge process, making it difficult to guarantee the stability and reliability of the reconstruction results. Summary of the Invention

[0005] Based on this, it is necessary to propose a three-dimensional dynamic reconstruction method for long-gap discharge channels to address the above problems.

[0006] A method for three-dimensional dynamic reconstruction of a long-gap discharge channel, the method comprising the following steps: By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at the given time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; The three-dimensional static models corresponding to multiple consecutive time points are combined in chronological order to generate the three-dimensional dynamic model of the discharge channel.

[0007] In the above scheme, the preprocessing of the group of synchronized frame images to obtain the corresponding time-aligned binarized images specifically includes: Each original frame image in the group of synchronized frame images is converted to grayscale to obtain a grayscale image; The grayscale image is binarized according to a preset grayscale threshold to obtain a set of binarized images corresponding to that time point; The frame with the largest number of white pixels in a set of binarized images corresponding to the time point is selected as the peak frame. The binarized image at the time point is then time-corrected based on the peak frame to obtain a time-aligned binarized image.

[0008] In the above scheme, the viewpoint correction of the time-aligned binarized image includes: Obtain the preset base matrix between at least two cameras that are synchronously triggered; The epipolar lines of the image are transformed into a perspective transformation matrix in the horizontal direction using the fundamental matrix. The time-aligned binarized image is resampled using the perspective transformation matrix to obtain an epipolar-aligned binarized image.

[0009] In the above scheme, the step of extracting white pixels at the edge of the discharge channel in the image after viewpoint correction to form the edge point set at that time point includes: Edge detection is performed on the aligned binarized image of the polar lines to extract the contour pixels of the discharge channel, which include the left boundary point and the right boundary point of the discharge channel. The left and right boundary points of the discharge channel are filtered to remove isolated noise points of the contour pixels; All the boundary points retained after filtering are combined to form the edge point set at that time point.

[0010] In the above scheme, the geometric constraint relationships between images from different viewpoints include:

[0011] Where p is the homogeneous pixel coordinate vector of the first view frame in the view-corrected image, p′ is the homogeneous pixel coordinate vector of the other view frames in the view-corrected image, and F is the preset fundamental matrix.

[0012] In the above scheme, an overdetermined system of equations is constructed based on the geometric constraint relationship between the edge point set and images from different viewpoints, specifically including: Using the geometric constraint relationship, select mutually matching boundary point pairs from the edge point set in images from different viewpoints. The boundary point pairs include mutually matching left boundary points or right boundary points. Establish a linear equation for each pair of the matched boundary points; Combine all the linear equations to construct the coefficient matrix A of the overdetermined system of equations.

[0013] In the above scheme, solving the overdetermined system of equations to obtain the three-dimensional spatial coordinate set corresponding to the edge point set specifically includes: The overdetermined system of equations constructed based on the edge point set is represented in matrix form: AQ=0 Where A is the coefficient matrix of the overdetermined system of equations, and Q is the homogeneous coordinate solution vector of the point in the three-dimensional space to be solved; Singular value decomposition of the coefficient matrix A of the overdetermined system of equations yields A = UΣV T Where U is the left singular vector matrix, Σ is a rectangular diagonal matrix with non-negative elements on its diagonal being singular values, and V is the right singular vector matrix. The column vector corresponding to the minimum singular value in the right singular vector matrix V is taken as the homogeneous coordinate solution vector Q of the point in the three-dimensional space to be solved. The homogeneous coordinate solution vector Q of the three-dimensional spatial point to be determined is homogeneously normalized to obtain the three-dimensional spatial coordinate set corresponding to the edge point set.

[0014] This application also proposes a three-dimensional dynamic reconstruction system for a long gap discharge channel, the system comprising: an image acquisition module, an edge point set acquisition module, a three-dimensional coordinate calculation module, and a three-dimensional model generation module; The image acquisition module acquires synchronous images of the long gap discharge process from different perspectives at multiple consecutive time points using at least two cameras deployed on the same plane and triggered synchronously, forming an original image sequence; wherein each time point corresponds to a set of synchronous frame images from all cameras. The edge point set acquisition module is used to perform the following processing on a set of synchronous frame images at each time point: preprocess the set of synchronous frame images to obtain the corresponding time-aligned binarized image; perform viewpoint correction on the time-aligned binarized image, and extract the white pixels at the edge of the discharge channel in the viewpoint-corrected image to form the edge point set at that time point. The three-dimensional coordinate calculation module is used to construct an overdetermined set of equations based on the edge point set at the time point and the geometric constraint relationship between images from different viewpoints; and to solve the overdetermined set of equations to obtain the three-dimensional spatial coordinate set corresponding to the edge point set at the time point. The three-dimensional model generation module is used to extract the cross-sectional edge points at different heights based on the three-dimensional spatial coordinate set at the time point, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate a three-dimensional static model corresponding to the time point; and combine the three-dimensional static models corresponding to multiple consecutive time points in chronological order to generate a three-dimensional dynamic model of the discharge channel.

[0015] This application also proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at the given time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; The three-dimensional static models corresponding to multiple consecutive time points are combined in chronological order to generate the three-dimensional dynamic model of the discharge channel.

[0016] This application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor in the following steps: By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at the given time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; The three-dimensional static models corresponding to multiple consecutive time points are combined in chronological order to generate the three-dimensional dynamic model of the discharge channel.

[0017] The embodiments of the present invention have the following beneficial effects: This invention utilizes at least two synchronously triggered cameras deployed on the same plane and executes a complete 3D reconstruction process for each time point's synchronized frame image group. Finally, by combining the 3D static models from all time points in sequence, it achieves 3D dynamic visualization of the entire transient evolution process of a long-gap discharge channel from generation to development to demise. Simultaneously, it employs a method based on geometric constraints and overdetermined equations to replace traditional gray-scale correlation matching, transforming the complex image matching problem into a mathematical optimal solution problem. This significantly reduces computational complexity and effectively overcomes plasma scintillation and noise interference, thereby significantly improving the accuracy of image matching. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] in: Figure 1 This is a schematic diagram of a three-dimensional dynamic reconstruction method for a long gap discharge channel in one embodiment. Detailed Implementation

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

[0021] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention; however, it will be apparent to those skilled in the art that the invention may be practiced without one or more of these details; in other instances, certain technical features well-known in the art have not been described in order to avoid confusion with the invention. It should be understood that the invention can be practiced in different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided to make the disclosure thorough and complete and to fully convey the scope of the invention to those skilled in the art.

[0022] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms, unless the context clearly indicates otherwise. The terms “comprising” and / or “including,” when used in this specification, identify the presence of said features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0023] To fully understand the present invention, a detailed structure will be presented in the following description in order to illustrate the technical solution proposed by the present invention; optional embodiments of the present invention are described in detail below, however, in addition to these detailed descriptions, the present invention may have other embodiments.

[0024] like Figure 1As shown, in one embodiment, a three-dimensional dynamic reconstruction method for a long gap discharge channel is provided. This method includes steps S101 to S103, which are detailed below: S101. By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. By deploying at least two synchronous cameras on the same plane, it is ensured that the physical phenomenon is captured at the same instant from different perspectives, eliminating reconstruction errors caused by shooting time differences; at the same time, the acquisition at multiple consecutive time points provides continuous time series data for subsequent construction of dynamic models, making it possible to capture the complete evolution process of the discharge channel.

[0025] Preferably, at least two cameras are used as two high-speed cameras that are triggered synchronously. Their specific parameters are preferably: a sampling frequency of not less than 1,000,000 frames / second (1 MHz) and a pixel resolution of not less than 1024×1024. Synchronous triggering is preferably achieved using an electromagnetic pulse signal generated during the discharge process as the trigger source to ensure that the initial moment of the discharge is captured.

[0026] S102. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and the white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at that time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; This step simplifies the data dimensions by using grayscale, and then uses binarization to effectively separate the luminescent area of ​​the discharge channel from the complex background noise by using a preset threshold, highlighting the target features. Furthermore, time alignment is used to ensure strict synchronization of different camera sequences on the time axis, providing a temporal consistency guarantee for subsequent accurate matching.

[0027] By utilizing perspective transformation and epipolar constraints for horizontal viewpoint correction, the complex two-dimensional image matching problem can be transformed into a simple one-dimensional search problem. This drastically reduces the computational cost of finding corresponding points in images from different viewpoints, while significantly improving the accuracy and efficiency of matching. Subsequently, edge points are extracted and formed into a point set, providing accurate and clean input data for three-dimensional computation.

[0028] By constructing an overdetermined system of equations using epipolar constraints and robustly solving it through singular value decomposition, the optimal 3D spatial coordinates can be stably calculated from noisy data. This method overcomes the shortcomings of traditional gray-scale matching methods, which are susceptible to illumination and flicker interference, and significantly improves the accuracy and reliability of 3D point cloud reconstruction.

[0029] Finally, a static model was constructed by calculating the center point and radius of the cross-section. This transformed the reconstructed result from an isolated point cloud into a continuous geometric model with clear physical meaning, representing the cylindrical morphology of the discharge channel. This significantly enhanced the ability to characterize the channel's physical properties, such as its thickness and orientation, providing a basis for quantitative analysis.

[0030] In some embodiments, the set of synchronized frame images is preprocessed to obtain corresponding time-aligned binarized images, specifically including: Each original frame image in the set of synchronized frame images is converted to grayscale to obtain a grayscale image; the grayscale image is then binarized according to a preset grayscale threshold to obtain a set of binarized images corresponding to that time point. The frame with the largest number of white pixels in a set of binarized images corresponding to the time point is selected as the peak frame. The binarized image at that time point is then time-corrected based on the peak frame to obtain a time-aligned binarized image.

[0031] Specifically, the number of white pixels in each group of binary images is counted, and the frame with the largest number is selected as the peak frame. This peak frame corresponds to the moment when the discharge channel emits the strongest light, ensuring a unified time reference. If the number of white pixels in the binary image of camera 1 at a certain time point is 8560 and that of camera 2 is 8552, then the frame of camera 1 is taken as the peak frame, and the timestamp of the image of camera 2 is adjusted through an interpolation algorithm to achieve time alignment between the two frames.

[0032] In some embodiments, viewpoint correction is performed on a time-aligned binarized image, including: Obtain the preset base matrix between at least two cameras that are synchronously triggered; The epipolar lines of the image are transformed into a perspective transformation matrix in the horizontal direction using the fundamental matrix; The time-aligned binarized image is resampled using a perspective transformation matrix to obtain an epipolar-aligned binarized image.

[0033] Specifically, the preset fundamental matrix F is invoked, and the perspective transformation matrix H is calculated through epipolar constraints to convert the epipolar lines of the binarized image of camera 2 into the horizontal direction.

[0034]

[0035]

[0036] Where H is the perspective transformation matrix, and F is the fundamental matrix. , These are the inverse matrix of the intrinsic parameters of camera 1 and the transpose inverse matrix of the intrinsic parameters of camera 2, respectively. Resample the temporally aligned binarized image of camera 2 to obtain an epipolar aligned binarized image. At this time, pixels of the same spatial point in the images of the two cameras are located in the same horizontal row.

[0037] In some embodiments, white pixels at the edges of the discharge channel are extracted from the viewpoint-corrected image to form an edge point set at that time point, including: Edge detection is performed on the aligned binarized image with polarity alignment to extract the contour pixels of the discharge channel, which include the left and right boundary points of the discharge channel. The left and right boundary points of the discharge channel are filtered to remove isolated noise points of the contour pixels; All the boundary points retained after filtering are combined to form the edge point set at that time point.

[0038] Specifically, the Canny edge detection algorithm is used, with a low threshold of 30 and a high threshold of 100 for binarized image processing aligned with the polarity, to extract the left and right boundary points of the discharge channel: For example, in the image from camera 1, the coordinates of the left boundary point are (156,230), (157,231)..., and the coordinates of the right boundary point are (210,230), (211,231)...; Perform 3×3 median filtering on the left and right boundary points respectively to remove isolated noise points and retain continuous boundary points; specifically, points that are more than 5 pixels away from their neighbors are removed. The filtered boundary points from both cameras are combined to form the edge point set for that time point: P = {p1, p2, ..., pN} (N is the total number of boundary points, in this embodiment N ≈ 1200), where each point pi contains the pixel coordinates (u1, v1) of camera 1 and the pixel coordinates (u2, v2) of camera 2.

[0039] In some embodiments, the geometric constraint relationships between images from different viewpoints include:

[0040] Where p is the homogeneous pixel coordinate vector of the first view frame in the view-corrected image, p′ is the homogeneous pixel coordinate vector of the other view frames in the view-corrected image, and F is the preset fundamental matrix.

[0041] In some embodiments, an overdetermined system of equations is constructed based on the geometric constraint relationships between edge point sets and images from different viewpoints, specifically including: Using geometric constraints, we can filter out matching boundary point pairs from the set of edge points in images from different viewpoints. Boundary point pairs include matching left or right boundary points. Establish a linear equation for each pair of matching boundary points; Combine all the linear equations to construct the coefficient matrix A of the overdetermined system of equations.

[0042] Specifically, p is the homogeneous coordinate of camera 1 (u1, v1, 1)T, and p′ is the homogeneous coordinate of camera 2 (u2, v2, 1). T ), Next, filter matching point pairs from the edge point set P and verify whether they match. For example, verify point p1=(156,230,1). T With p1′=(189,230,1) T Does p1 satisfy? T If Fp1′=0, then these two points are considered a matching pair.

[0043] For each pair of matching points, establish a linear equation: For example, p = K[R|t]Q; where R is the rotation matrix, t is the translation vector, and Q = (X,Y,Z,1). T Given three-dimensional homogeneous coordinates, it can be derived that: Expanding this equation, we get a linear equation in Q: Then, combine the linear equations of all matching point pairs to form the coefficient matrix A of the overdetermined system of equations.

[0044] In some embodiments, the three-dimensional spatial coordinates corresponding to the edge point set are obtained by solving the overdetermined system of equations. The collection specifically includes: The overdetermined system of equations constructed based on the edge point set is represented in matrix form: AQ=0 Where A is the coefficient matrix of the overdetermined system of equations, and Q is the homogeneous coordinate solution vector of the point in the three-dimensional space to be solved; singular value decomposition is performed on the coefficient matrix A of the overdetermined system of equations to obtain A=UΣV TWhere U is the left singular vector matrix, Σ is a rectangular diagonal matrix with non-negative elements on its diagonal being singular values, and V is the right singular vector matrix. Take the column vector corresponding to the minimum singular value in the right singular vector matrix V as the homogeneous coordinate solution vector Q of the point in the three-dimensional space to be solved; The homogeneous coordinate solution vector Q of the point to be solved in three-dimensional space is homogeneously normalized to obtain the three-dimensional coordinate set corresponding to the edge point set.

[0045] Specifically, singular value decomposition is performed on the coefficient matrix A of the overdetermined system of equations, yielding A = UΣV T ; Where Σ is a rectangular diagonal matrix, and the diagonal elements are singular values: : Find the minimum singular value in the right singular vector matrix V. Let the column vectors be Q, where Q = (X, Y, Z, 1). T By homogeneously normalizing Q, we obtain the three-dimensional spatial coordinates (X,Y,Z). For example, the three-dimensional coordinates of a certain boundary point are (0.82,1.50,2.35). Summarize the 3D coordinates of all matching point pairs to form a 3D spatial coordinate set for that point in time: .

[0046] Preferably, the cross-section is divided at 0.1m intervals along the Z-axis in the height direction, and the three-dimensional coordinate points within each cross-section are extracted as the edge points of the cross-section. For example, the edge points of the cross-section with Z=2.3m are (0.82,1.50,2.3), (0.85,1.52,2.3), (0.79,1.48,2.3)...; Calculate the center point and radius of the cross section: For each cross section, calculate the geometric midpoint of the edge points:

[0047] in, The X coordinate of the center point of the cross section is... The Y-coordinate of the center point of the cross section The Z-coordinate of the center point of the cross section It is the sum of the X coordinates of all edge points within this cross section. It is the sum of the Y coordinates of all edge points within this cross section. This represents the number of edge points within the cross section. Calculate the distance from the center point to each edge point, and take the average value as the cross-sectional radius; Then connect the sections in Z-axis order to form a cylindrical three-dimensional static model with the center point of the section as the axis and the radius of the section as the thickness.

[0048] S103. Combine the three-dimensional static models corresponding to multiple consecutive time points in chronological order to generate a three-dimensional dynamic model of the discharge channel.

[0049] By combining a series of static models at consecutive moments in chronological order, a three-dimensional dynamic evolution model was ultimately generated that can fully demonstrate the entire process of a long-gap discharge channel from its initiation and development to its breakdown. This breaks through the limitations of traditional two-dimensional imaging, achieving a complete record and visualization of the transient physical phenomenon of discharge with both high temporal resolution and true three-dimensional spatial information, providing an unprecedented observational tool for a deeper understanding of the discharge mechanism.

[0050] In summary, this invention achieves three-dimensional dynamic visualization of long-gap discharge channels, overcoming the spatial limitations of traditional two-dimensional imaging. By using polar constraints and solving overdetermined equations, two-dimensional matching is simplified to one-dimensional search, significantly improving computational efficiency while maintaining accuracy. The constructed continuous columnar model accurately depicts the channel's geometry, supporting quantitative analysis. By adjusting the binarization threshold, visualization analysis of ionization intensity distribution can be achieved. The entire method is robust, providing a universal three-dimensional dynamic observation solution for high-voltage experimental research.

[0051] This application also proposes a three-dimensional dynamic reconstruction system for a long gap discharge channel, the system including: an image acquisition module, an edge point set acquisition module, a three-dimensional coordinate calculation module, and a three-dimensional model generation module; The image acquisition module uses at least two cameras deployed on the same plane and triggered synchronously to acquire synchronous images of the long gap discharge process from different perspectives at multiple consecutive time points, forming an original image sequence; wherein each time point corresponds to a set of synchronous frame images from all cameras. The edge point set acquisition module is used to perform the following processing on a set of synchronization frame images at each time point: preprocess the set of synchronization frame images to obtain the corresponding time-aligned binarized image; perform viewpoint correction on the time-aligned binarized image, and extract the white pixels at the edge of the discharge channel in the viewpoint-corrected image to form the edge point set at that time point; The 3D coordinate calculation module is used to construct an overdetermined set of equations based on the edge point set at that time point and the geometric constraint relationship between images from different viewpoints; and to solve the overdetermined set of equations to obtain the 3D spatial coordinate set corresponding to the edge point set at that time point. The 3D model generation module is used to extract the cross-sectional edge points at different heights based on the 3D spatial coordinate set at the time point, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the 3D static model corresponding to the time point; combine the 3D static models corresponding to multiple consecutive time points in chronological order to generate the 3D dynamic model of the discharge channel.

[0052] This application also proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and the white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at that time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; The three-dimensional static models corresponding to multiple consecutive time points are combined in chronological order to generate a three-dimensional dynamic model of the discharge channel.

[0053] This application also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program, and the computer program is executed by the processor in the following steps: By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and the white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at that time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; The three-dimensional static models corresponding to multiple consecutive time points are combined in chronological order to generate a three-dimensional dynamic model of the discharge channel.

[0054] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by instructing related hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0055] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0056] The embodiments described above are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application's patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. The embodiments disclosed above are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made according to the claims of this invention are still within the scope of this invention.

Claims

1. A method for three-dimensional dynamic reconstruction of a long-gap discharge channel, characterized in that, The method includes: By using at least two cameras deployed on the same plane and triggered synchronously, synchronous images of the long gap discharge process from different perspectives are acquired at multiple consecutive time points to form an original image sequence; wherein, each time point corresponds to a set of synchronous frame images from all cameras. For a set of synchronized frame images at each time point, perform the following processing: The synchronized frame images are preprocessed to obtain the corresponding time-aligned binarized images; the time-aligned binarized images are then corrected by viewpoint, and white pixels at the edges of the discharge channels are extracted from the corrected images to form the edge point set at that time point. An overdetermined set of equations is constructed based on the edge point set at this time point and the geometric constraint relationship between images from different perspectives. Solving the overdetermined system of equations yields the three-dimensional spatial coordinate set corresponding to the edge point set at the given time point; Based on the three-dimensional spatial coordinate set at this time point, extract the cross-sectional edge points at different heights, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate the three-dimensional static model corresponding to this time point; The three-dimensional static models corresponding to multiple consecutive time points are combined in chronological order to generate the three-dimensional dynamic model of the discharge channel.

2. The three-dimensional dynamic reconstruction method for a long gap discharge channel according to claim 1, characterized in that, The preprocessing of the set of synchronized frame images to obtain the corresponding time-aligned binarized images specifically includes: Each original frame image in the group of synchronized frame images is converted to grayscale to obtain a grayscale image; The grayscale image is binarized according to a preset grayscale threshold to obtain a set of binarized images corresponding to that time point; The frame with the largest number of white pixels in a set of binarized images corresponding to the time point is selected as the peak frame. The binarized image at the time point is then time-corrected based on the peak frame to obtain a time-aligned binarized image.

3. The three-dimensional dynamic reconstruction method for a long gap discharge channel according to claim 2, characterized in that, The viewpoint correction of the time-aligned binarized image includes: Obtain the preset base matrix between at least two cameras that are synchronously triggered; The epipolar lines of the image are transformed into a perspective transformation matrix in the horizontal direction using the fundamental matrix. The time-aligned binarized image is resampled using the perspective transformation matrix to obtain an epipolar-aligned binarized image.

4. The three-dimensional dynamic reconstruction method for a long gap discharge channel according to claim 3, characterized in that, The step of extracting white pixels at the edges of the discharge channel in the image after viewpoint correction to form an edge point set at that time point includes: Edge detection is performed on the aligned binarized image of the polar lines to extract the contour pixels of the discharge channel, which include the left boundary point and the right boundary point of the discharge channel. The left and right boundary points of the discharge channel are filtered to remove isolated noise points of the contour pixels; All the boundary points retained after filtering are combined to form the edge point set at that time point.

5. The three-dimensional dynamic reconstruction method for a long gap discharge channel according to claim 1, characterized in that, The geometric constraints between images from different perspectives include: Where p is the homogeneous pixel coordinate vector of the first view frame in the view-corrected image, p′ is the homogeneous pixel coordinate vector of the other view frames in the view-corrected image, and F is the preset fundamental matrix.

6. The three-dimensional dynamic reconstruction method for a long gap discharge channel according to claim 5, characterized in that, An overdetermined system of equations is constructed based on the geometric constraints between the edge point set and images from different viewpoints, specifically including: Using the geometric constraint relationship, select mutually matching boundary point pairs from the edge point set in images from different viewpoints. The boundary point pairs include mutually matching left boundary points or right boundary points. Establish a linear equation for each pair of the matched boundary points; Combine all the linear equations to construct the coefficient matrix A of the overdetermined system of equations.

7. The three-dimensional dynamic reconstruction method for a long gap discharge channel according to claim 6, characterized in that, The process of solving the overdetermined system of equations to obtain the three-dimensional spatial coordinate set corresponding to the edge point set specifically includes: The overdetermined system of equations constructed based on the edge point set is represented in matrix form: AQ=0 Where A is the coefficient matrix of the overdetermined system of equations, and Q is the homogeneous coordinate solution vector of the point in the three-dimensional space to be solved; Singular value decomposition of the coefficient matrix A of the overdetermined system of equations yields A = UΣV T Where U is the left singular vector matrix, Σ is a rectangular diagonal matrix with non-negative elements on its diagonal being singular values, and V is the right singular vector matrix. The column vector corresponding to the minimum singular value in the right singular vector matrix V is taken as the homogeneous coordinate solution vector Q of the point in the three-dimensional space to be solved. The homogeneous coordinate solution vector Q of the three-dimensional spatial point to be determined is homogeneously normalized to obtain the three-dimensional spatial coordinate set corresponding to the edge point set.

8. A three-dimensional dynamic reconstruction system for a long-gap discharge channel, characterized in that, The system includes: an image acquisition module, an edge point set acquisition module, a three-dimensional coordinate calculation module, and a three-dimensional model generation module; The image acquisition module acquires synchronous images of the long gap discharge process from different perspectives at multiple consecutive time points using at least two cameras deployed on the same plane and triggered synchronously, forming an original image sequence; wherein each time point corresponds to a set of synchronous frame images from all cameras. The edge point set acquisition module is used to perform the following processing on a set of synchronous frame images at each time point: preprocess the set of synchronous frame images to obtain the corresponding time-aligned binarized image; perform viewpoint correction on the time-aligned binarized image, and extract the white pixels at the edge of the discharge channel in the viewpoint-corrected image to form the edge point set at that time point. The three-dimensional coordinate calculation module is used to construct an overdetermined set of equations based on the edge point set at the time point and the geometric constraint relationship between images from different viewpoints; and to solve the overdetermined set of equations to obtain the three-dimensional spatial coordinate set corresponding to the edge point set at the time point. The three-dimensional model generation module is used to extract the cross-sectional edge points at different heights based on the three-dimensional spatial coordinate set at the time point, take the geometric midpoint of the cross-sectional edge points as the cross-sectional center point, determine several cross-sectional radii based on the distance between the cross-sectional center point and the cross-sectional edge points, and generate a three-dimensional static model corresponding to the time point; and combine the three-dimensional static models corresponding to multiple consecutive time points in chronological order to generate a three-dimensional dynamic model of the discharge channel.

9. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the processor performs the steps of the method as described in any one of claims 1 to 7.

10. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images