A method and related apparatus for three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras

Abstract

This invention provides a method and related apparatus for the three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras. The method includes determining the shooting points of multiple high-speed cameras centered on the discharge gap; capturing dynamic images of the long air gap discharge at each shooting point using the high-speed cameras; processing the images captured at each shooting point to identify the discharge channels in the images and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images; determining the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap; matching the discharge channel images captured at different shooting points and at different times for each discharge channel, and constructing the discharge channels in three-dimensional space at different times; and constructing a complete dynamic three-dimensional discharge channel based on the discharge channels in three-dimensional space at different times. This invention achieves efficient and accurate three-dimensional reconstruction by utilizing the positional relationship between multiple high-speed cameras and the discharge channels.

Description

Technical Field

[0001] This invention belongs to the field of gas discharge observation and analysis technology, specifically relating to a dynamic discharge channel three-dimensional adaptive reconstruction method and related device based on multiple high-speed cameras. Background Technology

[0002] With rapid socio-economic development and continuous acceleration of urbanization, energy demand is increasing year by year, prompting the continuous expansion of the power transmission and distribution network. The mismatch between the distribution of energy resources and consumption areas in my country necessitates long-distance power transmission for achieving balanced energy allocation. Ultra-high voltage (UHV) transmission lines, with their high voltage levels and low losses, can effectively reduce energy losses and environmental pollution, meet domestic energy demand, and promote the optimal allocation of national power resources. Under high-voltage and high-current operating conditions, the external insulation design of UHV transmission systems is crucial, and the discharge characteristics of long air gaps, as the primary form of external insulation, have become the core and key direction for the external insulation design of UHV transmission systems.

[0003] Current observations of long air gap discharge characteristics in the field of gas discharge are mostly limited to two-dimensional observations, making it difficult to fully present the three-dimensional morphology and dynamic evolution of the discharge channel. Although the development of high-speed photography technology has made it feasible to simultaneously capture discharge phenomena from different angles using multiple high-speed cameras, thus providing a feasible way to reconstruct the discharge channel in three dimensions, dynamic discharge channels have characteristics such as rapid changes, uneven light intensity, and complex morphology, posing a significant challenge to traditional feature point detection and tracking algorithms. When existing multi-view geometry principles are applied to the three-dimensional reconstruction of dynamic discharge channels, problems such as easy loss of feature points and unstable tracking often occur, resulting in a significant reduction in the accuracy and reliability of the three-dimensional reconstruction. How to achieve efficient and accurate three-dimensional reconstruction remains a hot and difficult problem that urgently needs to be solved in the current research field. Summary of the Invention

[0004] In view of this, the present invention provides a method and related device for dynamic three-dimensional adaptive reconstruction of discharge channels based on multiple high-speed cameras, aiming to achieve high-precision three-dimensional reconstruction of discharge channels by optimizing the shooting layout and data processing algorithm, and providing a new technical means for the study of gas discharge phenomena.

[0005] To achieve the above objectives, the technical solution provided by the present invention is as follows:

[0006] In a first aspect, the present invention provides a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras, comprising the following steps:

[0007] Using the discharge gap as the center, determine the shooting points of multiple high-speed cameras;

[0008] At each shooting point, dynamic images of long air gap discharges are captured using high-speed cameras.

[0009] The images captured at each shooting point are processed to identify the discharge channels in the images and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images;

[0010] Determine the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap;

[0011] The discharge channel images captured at different shooting points and at different times are matched, and the discharge channel in three-dimensional space at different times is constructed based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image.

[0012] A complete dynamic three-dimensional discharge channel is constructed based on the discharge channels of each discharge channel in three-dimensional space at different times.

[0013] Furthermore, the images captured at each shooting point are processed to identify the discharge channels in the images and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images, including:

[0014] The images captured at each shooting point are converted into grayscale images;

[0015] The grayscale image is processed using a grayscale image processing method based on morphological filtering to obtain the upper and lower envelope surfaces of the grayscale image;

[0016] The upper and lower envelope surfaces are smoothed by using a mean filter to obtain smooth upper and lower envelope surfaces;

[0017] The characteristics of the discharge channel are identified based on the smooth upper and lower envelope surfaces, and the grayscale image is processed into a binary image while preserving the characteristics.

[0018] Discharge channels are identified in binary images, and a two-dimensional coordinate sequence of the discharge channels in the image is constructed.

[0019] Furthermore, the upper and lower envelope surfaces are determined according to the following formula:

[0020]

[0021]

[0022] In the formula, and Pixel positions The upper and lower envelope surface values ​​at the location; In pixel position Centered The collection of pixels in a window; Original grayscale image At pixel position The gray value at that location; ⊕ represents the morphological dilation filter; is a morphological erosion filter, and b is a binary indicator function.

[0023] Furthermore, the smooth upper and lower envelope surfaces are determined according to the following formula:

[0024]

[0025]

[0026] In the formula, and Pixel positions The smoothed upper and lower envelope surface values; and Pixel positions The upper and lower envelope surface values ​​at the location.

[0027] Furthermore, the images of each discharge channel captured at different shooting points are matched, including:

[0028] Extract a point from the path of the discharge channel as a reference point and obtain the two-dimensional coordinates of the reference point in each discharge channel image; the two-dimensional coordinates include horizontal and vertical coordinates.

[0029] Based on the three-dimensional coordinates of the shooting point and the two-dimensional coordinates of the reference point corresponding to the discharge channel image, calculate the azimuth and elevation angles of the reference point observed at the shooting point.

[0030] In the same three-dimensional coordinate system, any two shooting points are taken and observed along the perpendicular line to the common reference point according to their corresponding azimuth and elevation angles, thus obtaining two observation points on the perpendicular line. The common reference point is the point determined by projecting the two horizontal coordinates of the reference point onto the horizontal plane in the discharge channel image taken by the two shooting points.

[0031] The distance difference between the two observation points is used to determine whether the images of the discharge channel captured by the two shooting points match. If they match, the image is assigned to the image set of the corresponding discharge channel. If not, the shooting point is changed until all shooting points have been traversed.

[0032] Furthermore, when determining whether the discharge channel images captured by the two shooting points match based on the distance difference between the two observation points, the distance difference between the two observation points is converted into an elevation angle error before the judgment is made. The elevation angle error is calculated according to the following formula:

[0033]

[0034] In the formula, This refers to the elevation angle error; It represents the distance difference; First shooting point Distance from common reference point Perpendicular line Horizontal distance, First shooting point The horizontal plane and the vertical line The intersection; Second shooting location Distance from common reference point Perpendicular line Horizontal distance, Second shooting location The horizontal plane and the vertical line The intersection;

[0035] The distance difference is determined by the following formula:

[0036]

[0037]

[0038]

[0039]

[0040]

[0041] In the formula, and The first shooting point According to the corresponding azimuth and elevation angles Second shooting point According to the corresponding azimuth and elevation angles on the vertical line The two observation points observed above; and They are common reference points Distance between two observation points and The distance; and These are the distances between the two intersection points and the two observation points, respectively; and The first shooting point Second shooting point Height; and These are the distances between the points projected from the two shooting points onto the horizontal plane where the common reference point is located, and the common reference point itself.

[0042] Furthermore, based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image, discharge channels in three-dimensional space at different times are constructed, including:

[0043] For any reference point in the discharge channel Using the three-dimensional coordinates of at least two shooting points matched by the reference point, the distances from the two shooting points to the corresponding two observation points are calculated. and ;

[0044] Based on distance and Calculate reference point The distances to the two observation points are as follows:

[0045]

[0046]

[0047] In the formula, and Reference points Distance to the two observation points;

[0048] For reference point Its three-dimensional reconstruction coordinates are determined according to the following formula:

[0049]

[0050]

[0051]

[0052] In the formula, , and For 3D reconstruction coordinates; and The reference points are defined in the discharge channel images taken from the common reference point to the first and second shooting points, respectively. Horizontal coordinates;

[0053] The three-dimensional reconstructed coordinates of all reference points are combined to form the three-dimensional reconstructed coordinates of the discharge channel.

[0054] Secondly, the present invention provides a three-dimensional adaptive reconstruction device for dynamic discharge channels based on multiple high-speed cameras, comprising:

[0055] The shooting point location determination module is used to determine the shooting points of multiple high-speed cameras with the discharge gap as the center.

[0056] The image acquisition module is used to capture dynamic images of long air gap discharges at various shooting points using a high-speed camera.

[0057] The image recognition module is used to process the images captured at each shooting point, identify the discharge channels in the images, and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images;

[0058] The three-dimensional coordinate determination module is used to determine the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap;

[0059] The image matching and processing module is used to match the images of each discharge channel captured at different shooting points and at different times, and to construct the discharge channel in three-dimensional space at different times based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image.

[0060] The 3D reconstruction module is used to construct a complete dynamic 3D discharge channel based on the 3D discharge channel of each discharge channel at different times.

[0061] Thirdly, the present invention provides a computer device, the device including a processor and a memory:

[0062] The memory is used to store computer programs and send the instructions of the computer programs to the processor;

[0063] The processor executes instructions from a computer program, such as the first aspect of a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras.

[0064] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras, as described in the first aspect.

[0065] In summary, this invention provides a method and related apparatus for the three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras. The method includes determining the shooting points of multiple high-speed cameras centered on the discharge gap; capturing dynamic images of the long air gap discharge at each shooting point using the high-speed cameras; processing the images captured at each shooting point to identify the discharge channels in the images and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images; determining the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap; matching the discharge channel images captured at different shooting points and at different times for each discharge channel, and constructing the discharge channel in three-dimensional space at different times based on the two-dimensional coordinate sequence of each discharge channel in the matched images and the three-dimensional coordinate position of the shooting point corresponding to the matched images; and constructing a complete dynamic three-dimensional discharge channel based on the discharge channel in three-dimensional space at different times. This invention achieves efficient and accurate three-dimensional reconstruction by utilizing the positional relationship between multiple high-speed cameras and the discharge channels. Attached Figure Description

[0066] 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.

[0067] Figure 1 A flowchart illustrating a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras, provided in an embodiment of the present invention;

[0068] Figure 2 This is a schematic diagram of the camera arrangement for a high-speed camera provided in an embodiment of the present invention;

[0069] Figure 3 This is a schematic diagram of the discharge channel image before and after preprocessing provided in an embodiment of the present invention;

[0070] Figure 4 These are front and side views of the discharge channel image after image enhancement processing provided in an embodiment of the present invention;

[0071] Figure 5 This is a schematic diagram illustrating the relative positional relationship between the shooting point and the discharge channel provided in an embodiment of the present invention.

[0072] Figure 6 A reconstructed three-dimensional dynamic development path diagram of the leader discharge provided in an embodiment of the present invention;

[0073] Figure 7This is a block diagram of a dynamic discharge channel three-dimensional adaptive reconstruction device based on multiple high-speed cameras, provided in an embodiment of the present invention.

[0074] Figure 8 This is a block diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0075] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0076] Please see Figure 1 This application provides a method for three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras, including the following steps:

[0077] S1: Determine the shooting points of multiple high-speed cameras with the discharge gap as the center.

[0078] It should be noted that this step prepares for subsequent comprehensive and multi-angle filming of the long air gap discharge process. The proper determination of the shooting point is the basis for obtaining accurate three-dimensional information of the discharge channel.

[0079] The determined shooting points need to take into account factors such as the shape and size of the discharge area. For different types of discharge test objects (such as rod-plate, ring-plate, etc.), the location of the corona initiation point and the area where discharge may occur are different. The location and number of shooting points should be determined according to the specific situation to ensure that the entire area where discharge may occur can be covered.

[0080] S2: At each shooting point, a high-speed camera captures dynamic images of discharge over a long air gap.

[0081] It should be noted that this step can capture continuous dynamic images of the discharge process in a long air gap, providing raw data for subsequent image processing and 3D reconstruction.

[0082] During this step, high-speed cameras need to be installed at each designated shooting point, and parameters such as frame rate, exposure time, and resolution need to be adjusted to accommodate the rapid changes in the discharge process. When the discharge occurs, all high-speed cameras should be activated simultaneously to ensure the complete dynamic process of the discharge is recorded. During shooting, the shooting environment should be kept as stable as possible to avoid external factors (such as vibration, changes in lighting, etc.) affecting image quality.

[0083] S3: Process the images captured at each shooting point, identify the discharge channels in the images, and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images.

[0084] It should be noted that this step can extract the information of the discharge channel from the original image and convert it into a two-dimensional coordinate form that is convenient for subsequent three-dimensional reconstruction.

[0085] S4: Determine the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap.

[0086] It should be noted that this step can establish the positional relationship of all shooting points in a unified three-dimensional space, so as to facilitate subsequent three-dimensional reconstruction based on multi-view images.

[0087] During implementation, a three-dimensional spatial coordinate system is first established with the center of the discharge gap as the origin. For each shooting point, its coordinates (x, y, z) in this three-dimensional coordinate system can be measured using high-precision measuring equipment (such as a total station, laser rangefinder, etc.). These coordinates will be used to determine the spatial geometric relationship between images of the same discharge channel taken from different shooting points.

[0088] S5: Match the discharge channel images captured at different shooting points and at different times for each discharge channel, and construct the discharge channel in three-dimensional space at different times based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image.

[0089] It should be noted that this step accurately selects images belonging to the same discharge channel from a large number of discharge channel images taken at many different shooting points and at different times, and constructs the three-dimensional spatial morphology of the discharge channel at different times based on these matching images and related coordinate information, gradually restoring the dynamic evolution process of the discharge channel.

[0090] Specifically, the first step is to analyze the geometric relationships (such as angles and distances) of different shooting points when capturing the same discharge channel, as well as potential error factors (such as camera installation deviations and measurement errors). Based on these factors, a suitable image matching algorithm (such as a feature point matching algorithm) is used to match images belonging to the same discharge channel by analyzing the positional relationships of feature points (such as specific points on the discharge channel) in different images. For example, it can be determined whether they come from the same discharge channel by comparing the relative positions and grayscale values ​​of feature points in different images.

[0091] Furthermore, considering images captured at different times, time factors also need to be processed. It may be necessary to group images captured at the same or similar times together for matching, based on information such as timestamps or frame rates, to ensure matching accuracy.

[0092] After image matching is completed, for each successfully matched discharge channel, by combining its two-dimensional coordinate sequence in the matched image and the three-dimensional coordinate position of the corresponding shooting point in the matched image, the two-dimensional coordinates can be converted into three-dimensional coordinates using triangulation principles or other three-dimensional reconstruction algorithms. Specifically, given the three-dimensional coordinates of two or more shooting points and the two-dimensional coordinates of the discharge channel in the images captured at those points, the three-dimensional coordinates of the corresponding points on the discharge channel can be calculated using geometric relationships. By performing this processing on the images of each discharge channel at different times, the three-dimensional spatial morphology of the discharge channel at different times can be constructed.

[0093] S6: Construct a complete dynamic three-dimensional discharge channel based on the discharge channels of each discharge channel in three-dimensional space at different times.

[0094] It should be noted that this step integrates the three-dimensional spatial morphology of each discharge channel at different times to construct a complete three-dimensional model that can reflect the entire dynamic evolution process of the discharge channel, so as to study the discharge phenomenon more comprehensively and intuitively.

[0095] Specifically, based on the three-dimensional spatial morphology of each discharge channel constructed in the previous steps at different times, they are connected in chronological order. During the connection process, it is important to maintain the continuity of the discharge channels in both time and space, avoiding jumps or discontinuities. Appropriate interpolation algorithms (such as linear interpolation, spline interpolation, etc.) can be used to process the three-dimensional channels at adjacent times to fill in any possible time intervals, making the entire dynamic three-dimensional discharge channel smoother and more natural. Furthermore, the constructed dynamic three-dimensional discharge channels can be further processed according to actual needs, such as rendering and annotation, to better display and analyze the discharge phenomenon.

[0096] This embodiment provides a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras. It determines the shooting points of multiple high-speed cameras centered on the discharge gap, captures dynamic images of the long air gap discharge at these points, processes the images to identify the discharge channel and construct a two-dimensional coordinate sequence, and simultaneously determines the three-dimensional coordinate positions of the shooting points. Based on the three-dimensional coordinates of the shooting points and the error relationship matching images of the same discharge channel captured from different shooting points, the three-dimensional reconstruction of the discharge channel is finally performed by combining the two-dimensional coordinate sequence of the discharge channel and the three-dimensional coordinates of the shooting points. This method has several advantages. In terms of high precision, through reasonable layout of shooting points and precise calibration, combined with advanced image processing algorithms, it can accurately capture the three-dimensional morphology of the discharge channel at different times. In terms of high efficiency, the simultaneous shooting by multiple high-speed cameras, combined with efficient image processing and three-dimensional reconstruction algorithms, greatly shortens the data processing time. In terms of comprehensiveness and dynamism, compared with traditional two-dimensional observation, it can record the discharge process from all directions and multiple angles and construct a three-dimensional model, and can also capture the dynamic evolution of the discharge channel, providing reliable and intuitive comprehensive data support for the study of dielectric breakdown phenomena.

[0097] In one embodiment, step S1 can be achieved through the following steps:

[0098] (1) Site survey and planning: First, the discharge experiment site was surveyed to determine the specific location, size and shape of the discharge gap. At the same time, the environmental conditions of the site were observed, including lighting conditions and possible sources of interference, to provide a basis for the subsequent layout design of the shooting points.

[0099] (2) Shooting point layout design: The specific discharge conditions of different samples should be fully considered. Centered on the discharge gap, a reasonable shooting point layout should be designed according to the size and shape of the discharge area. Ensure coverage of the discharge area from multiple angles (e.g., front and back, left and right, top and bottom) to obtain sufficient three-dimensional information. For example... Figure 2 As shown, if the sample is a rod or a ball, considering that the halo point of this type of gap is at the top of the electrode, only two high-speed cameras need to shoot orthogonally. If the sample is a ring plate, a ring combination gap, or a double ring combination gap, considering that the halo point of the ring may be anywhere between the horizontal and vertical tangent points below the ring, and that it will tilt due to wind force, high-speed cameras need to be arranged in all four horizontal directions, and the cameras in the four directions should have a certain height difference to ensure that the discharge area is covered from multiple angles in order to obtain sufficient three-dimensional information.

[0100] (3) Shooting Point Calibration: Markers are placed at each shooting point, or high-precision measuring equipment is used to determine the precise coordinates of each shooting point in three-dimensional space. A specially designed marker with known dimensions and shape is placed at each shooting point. Then, a total station or laser rangefinder is used to measure each marker to determine its coordinates in three-dimensional space. These coordinates are recorded and input into the data processing and analysis system for subsequent three-dimensional reconstruction.

[0101] In one embodiment, step S2 can be implemented by the following steps:

[0102] (1) Equipment Preparation and Debugging: Inspect and debug to ensure all high-speed cameras are working properly. First, perform a visual inspection of each high-speed camera to check for any damaged or loose parts. Then, connect the power supply and data cable, turn on the power switch of the high-speed camera, enter the settings interface, and adjust parameters such as the viewfinder, lens focal length, and exposure time to adapt to the shooting requirements of the discharge process. By adjusting various parameters, ensure that the moment of discharge can be clearly captured while ensuring the highest frame rate during the shooting process.

[0103] (2) Synchronization Trigger Setting: A synchronization control device is used to ensure that all high-speed cameras start shooting at the same time to guarantee the temporal consistency of the image sequence. A synchronization controller is used, and all high-speed cameras are connected to the synchronization controller. The trigger signal source of the synchronization controller is set, which can be an external trigger signal (such as a pulse generator) or an internal trigger signal (such as a timer). When the trigger signal is issued, the synchronization controller simultaneously sends the trigger signal to all high-speed cameras, so that they start shooting at the same time.

[0104] (3) Shooting Execution: Start shooting and record a continuous sequence of images during the discharge process. Maintain a stable shooting environment and avoid external interference affecting image quality. Shock absorption devices can be used to reduce the impact of ground vibrations on the high-speed camera. Simultaneously, turn off unnecessary light sources on site to avoid light interference. If possible, use a lens hood to further reduce the influence of external light.

[0105] In one embodiment, step S3 can be implemented through the following steps:

[0106] (1) Image preprocessing: Denoising and contrast enhancement are performed on the images captured by each camera to improve image quality. Median or Gaussian filters are used to denoise the images and remove noise points. Histogram equalization or contrast stretching are used to enhance the contrast of the images, making the discharge channels more clearly visible in the images.

[0107] (2) Discharge Channel Identification: Image processing algorithms (such as edge detection, threshold segmentation, etc.) are used to identify the discharge channels at different times and extract their contours. The Canny edge detection algorithm can be used to perform edge detection on the image to find the edges of the discharge channels. Then, the threshold segmentation algorithm is used to separate the discharge channels from the background and extract their contours.

[0108] Furthermore, morphological algorithms can be used to identify discharge channels in order to construct a two-dimensional coordinate sequence of the discharge channels in the corresponding image. The process of identifying discharge channels is as follows:

[0109] 1) Identifying discharge channels using morphological algorithms

[0110] The formula for calculating the upper and lower envelope surfaces of a W×H grayscale image I based on morphological filtering using BEMD is as follows:

[0111]

[0112]

[0113] Where ⊕ represents a morphological dilation filter. Let Zxy represent the morphological erosion filter, Zxy represent the set of pixels in a w×w window centered at pixel (x, y), b is the binary indicator function on Zxy, and w is the average extremum distance of the grayscale image I, calculated by the following formula:

[0114]

[0115] Where N represents the average number of local maxima and minima in image I.

[0116] 2) By smoothing the upper and lower envelopes using the following mean filter, a smoother envelope can be obtained.

[0117]

[0118]

[0119] By directly determining the envelope surface in BEMD computation, a multi-level spatial feature decomposition of an image, from detail to the whole, can be obtained. Similar to BEMD based on order statistics, this method reduces the problems of over-dilation or under-dilation that may occur during envelope construction, effectively capturing image features at various scales. Importantly, the morphological and mean filtering techniques it uses can be executed quickly, resulting in high computational efficiency.

[0120] 3) For the original image (i.e. Figure 3(a) Preprocessing is performed by reading the image as a two-dimensional grayscale array, with grayscale values ​​ranging from [0, 255]. The preprocessed image is shown below. Figure 3 As shown. Next, image enhancement is performed on the photo. The array matrix is ​​scanned row by row to identify the positions with a grayscale value of 255, thus identifying the location of the discharge channel. Then, with the discharge start position as the center and the coordinates set to (5, 5, 10), and the discharge end position as (x, y, 0), the image is cropped proportionally. The enhanced discharge channel image is shown below. Figure 4 As shown.

[0121] (3) Two-dimensional coordinate construction: Based on the image coordinate system and camera parameters (such as focal length, pixel size, etc.), the outline of the discharge channel is converted into a two-dimensional coordinate sequence. The origin and coordinate axis directions of the image coordinate system are determined, and the actual physical size corresponding to each pixel in the image is calculated based on the camera's focal length and pixel size. Finally, the coordinates of each point on the outline of the discharge channel in the image coordinate system are converted into actual physical coordinates to obtain the two-dimensional coordinate sequence of the discharge channel.

[0122] In one embodiment, step S4 can be implemented by the following steps:

[0123] (1) Establish a three-dimensional coordinate system: A three-dimensional rectangular coordinate system is established with the geometric center of the discharge gap as the origin. The direction and scale of the coordinate axes are determined based on the results of the on-site investigation and the calibration of the shooting points.

[0124] (2) Camera position calibration: Input the coordinates of each shooting point in the three-dimensional coordinate system into the system.

[0125] In one embodiment, step S5 includes matching images from different shooting points and matching images from different times.

[0126] Matching images from different times can be achieved through the following steps:

[0127] (1) Adaptive Threshold Setting: The threshold for feature point detection is dynamically adjusted based on the light intensity distribution characteristics of the discharge channel. Local contrast analysis is used to achieve adaptive recognition of regions with different light intensities. First, histogram analysis or local contrast calculation is used to analyze the light intensity distribution of the discharge channel. Then, the threshold for feature point detection is dynamically adjusted based on the characteristics of the light intensity distribution. If the light intensity of the discharge channel is high, the threshold can be appropriately lowered to improve the sensitivity of feature point detection; if the light intensity is low, the threshold can be appropriately increased to reduce false detections.

[0128] (2) Pre-screening of optical feature points: Using morphological filtering or edge detection algorithms, possible bright areas and edges on the discharge channel are preliminarily screened as candidate feature points;

[0129] (3) Quality assessment of optical feature points: A feature point quality assessment mechanism is used to comprehensively consider factors such as the stability, significance and uniformity of distribution of feature points, and select high-quality feature points from the candidate feature points;

[0130] The stability of feature points is evaluated by analyzing feature point matching across multiple frames. Stable feature points should maintain a high matching rate across different frames. After detecting feature points across multiple frames and obtaining candidate feature points in different frames, nearest neighbor matching or Random Sample Consensus (RANSAC) matching is used to match the feature points in different frames. The matching rate of each feature point in different frames is calculated, and feature points with higher matching rates are selected as stable feature points.

[0131] The saliency of a feature point is evaluated by considering factors such as brightness and contrast in the region where the feature point is located. Regions with high brightness and contrast typically contain more valuable feature points. Local contrast calculations or brightness statistics are used to analyze the brightness and contrast of the region where the feature point is located. The saliency of the feature point is evaluated based on its brightness and contrast. Feature points with higher brightness and contrast are selected as salient feature points.

[0132] Ensure that the selected feature points are evenly distributed along the discharge channel, avoiding excessive density in some areas and sparseness in others. Use spatial distribution statistics or cluster analysis to calculate the positional distribution of each feature point along the discharge channel. Evaluate the uniformity of feature point distribution based on the positional distribution. If feature points are too dense in some areas, some redundant feature points can be deleted; if feature points are too sparse in some areas, some new feature points can be added to ensure that feature points are evenly distributed along the discharge channel.

[0133] (4) Dynamic tracking of optical feature points: Based on the Kalman filter tracking method and combined with the motion characteristics of the discharge channel, the selected feature points are tracked stably. During the tracking process, the feature point descriptor is updated in real time to adapt to the dynamic changes of the discharge channel.

[0134] The Kalman filter is used to predict the location of feature points in the next frame, and this prediction is combined with the actual detected location for updating. This helps maintain tracking stability when the image is blurred or occluded. During tracking, the descriptors of the feature points (such as SIFT, SURF, or ORB descriptors) are updated in real time to reflect the dynamic changes in the discharge channel. Efficient matching algorithms such as FLANN (Fast Nearest Neighbor Search Library) are used to match feature points between adjacent frames.

[0135] A Kalman filter is used to predict the position of feature points in the next frame, and this prediction is updated based on the actual detected position. This helps maintain tracking stability when the image is blurred or occluded. First, the state equation and observation equation of the Kalman filter are established to describe the motion state and observation model of the feature points. Then, based on the position and velocity of the feature points in the current frame, the position of the feature points in the next frame is predicted. When the next frame image is actually detected, the predicted position is compared with the actual detected position, and the state estimate of the Kalman filter is updated based on the difference. Through continuous prediction and updating, stable tracking of feature points can be achieved.

[0136] During tracking, the descriptors of feature points (such as SIFT, SURF, or ORB descriptors) are updated in real time to reflect the dynamic changes in the discharge channel. An appropriate feature point descriptor, such as the SIFT descriptor, is selected. During tracking, the feature point descriptors are updated in real time based on the dynamic changes in the discharge channel. If the shape of the discharge channel changes, the SIFT descriptors of the feature points can be recalculated to reflect the new shape features.

[0137] Efficient matching algorithms such as FLANN (Fast Nearest Neighbor Search) are employed to match feature points between adjacent frames. The FLANN algorithm is used to match feature points in adjacent frames. First, a database of feature point descriptors is established. Then, the FLANN algorithm is used to quickly search the database for the feature point most similar to the feature point in the current frame. By matching feature points, the correspondence between feature points in different frames can be determined, thereby enabling feature point tracking and matching of discharge channel images at different times.

[0138] The following process is used for matching images from different shooting points:

[0139] (1) Extract a point from the path of the discharge channel as a reference point and obtain the two-dimensional coordinates of the reference point in each discharge channel image; the two-dimensional coordinates include horizontal coordinates and vertical coordinates.

[0140] First, the reference position must be determined to measure the elevation and azimuth angles. In field experiments, the camera's optical axis is difficult to measure, and the horizontal direction of the image is not always parallel to the ground; therefore, the image angular coordinates must be initialized. The vertex can be used as a reference point for rotating the image to calibrate the horizontal direction. Using the observation point and a known geographical reference point, the elevation and azimuth angles can be calculated, revealing the actual differences. Often, the measured angle differences in the image differ from the actual values. Image rotation and the least squares method can bring these differences closer to the actual values, establishing a coordinate transformation relationship. After rotation, the vertex of a building in the image is selected as the reference point; in actual space, this reference point has an azimuth angle α relative to the observation station. real and elevation angle β real The reference point in the image is α relative to the center point. image and βimage The angular coordinates of a single pixel can be calculated. An initial coordinate sequence can be obtained by processing two observation stations.

[0141] like Figure 5 As shown, let A be the main observation station, located directly above the X-axis. A' is the projection point of the main observation station on the Z=0 plane, and h1 represents the height of the camera at the main observation station. A set of azimuth and elevation coordinates is selected from the lightning channel captured by the main observation station, denoted as α1 and β1. B is the secondary observation station, located directly above the Y-axis. B' is the projection point of the secondary observation station on the Z=0 plane, and h2 represents the height of the camera at the secondary observation station. A set of azimuth and elevation coordinates is selected from the lightning channel captured by the secondary observation station, denoted as α2 and β2. In the Z=0 plane, a ray with vertex A' and azimuth angle α1 intersects a ray with vertex B' and azimuth angle α2 at point P. In the Z=h1 plane, a ray with vertex A and azimuth angle α1 intersects a ray with vertex P that is perpendicular to the Z=0 plane at point P1. In the Z=h2 plane, a ray with vertex B and azimuth angle α2 intersects a perpendicular line with vertex P at point P2.

[0142] (2) Based on the three-dimensional coordinates of the shooting point corresponding to the discharge channel image and the two-dimensional coordinates of the reference point, calculate the azimuth and elevation angles of the reference point observed at the shooting point, i.e. and .

[0143] (3) Under the same three-dimensional coordinate system, take any two shooting points and observe them along the perpendicular line of the common reference point according to the corresponding azimuth and elevation angles to obtain two observation points on the perpendicular line of the two shooting points; the common reference point is the point determined by projecting the two horizontal coordinates of the reference point onto the horizontal plane in the discharge channel image taken by the two shooting points.

[0144] according to Figure 5 It can be seen that there is , , ,and:

[0145]

[0146]

[0147]

[0148] , , ,and:

[0149]

[0150]

[0151]

[0152] The horizontal distance from the camera at the main observation station to the origin of the coordinate system. This represents the horizontal distance from the camera at the secondary observation station to the origin of the coordinate system. (In a right angle...) In the middle, it can be obtained through calculation. , and .

[0153]

[0154]

[0155]

[0156] exist In the middle, using the law of sines, we can obtain:

[0157]

[0158]

[0159] With vertex A, a ray with azimuth and elevation angles of α1 and β1 intersects the perpendicular PL at point L1; with vertex B, a ray with azimuth and elevation angles of α2 and β2 intersects the perpendicular PL at point L2.

[0160] (4) Determine whether the discharge channel images captured by the two shooting points match based on the distance difference between the two observation points. If they match, they are assigned to the image set of the corresponding discharge channel. If not, the shooting point is changed until all shooting points have been traversed.

[0161] The height error (i.e., distance difference) between L1 and L2 is as follows:

[0162]

[0163] In a further embodiment, the distance difference can be converted into elevation angle errors of AL1 and BL2 before matching and judgment, as follows:

[0164]

[0165] The above calculation process allows us to determine the elevation difference between any set of lightning channel angle coordinates obtained from a specific station and the entire lightning channel angle coordinates obtained from another observation station. The two sets of angle coordinates with the smallest values ​​are highly likely to meet the matching criteria. Typically, in actual lightning channel reconstruction, the minimum elevation difference (Δβ) is...min The value is not zero. δ is defined as the acceptable error range. When Δβ min When Δβ is less than δ, it means that a matching angle combination has been found; while when Δβ is less than δ, it means that a matching angle combination has been found. min When the value is greater than δ, it can be determined that there is no coordinate sequence in the image of the secondary observation station that matches the same set of angles as the primary observation station.

[0166] By now, the matching of images from different times and shooting points can be completed using the above method. The remaining part of step S5 can be achieved through the following steps:

[0167] (5) Three-dimensional coordinate transformation: Based on the camera's intrinsic and extrinsic parameters (such as rotation matrix, translation vector, etc.) and the feature point matching results, the two-dimensional coordinate sequence is transformed into a three-dimensional coordinate sequence;

[0168] First, the camera is calibrated to obtain its intrinsic and extrinsic parameters (rotation matrix, translation vector, focal length, etc.). Using the camera's intrinsic and extrinsic parameters and the two-dimensional coordinates of the feature points, the three-dimensional coordinates of the feature points are calculated through the inverse process of perspective projection. Considering camera calibration errors and image processing errors, the calculated three-dimensional coordinates are then corrected and filtered as necessary.

[0169] First, the camera is calibrated to obtain its intrinsic and extrinsic parameters (rotation matrix, translation vector, focal length, etc.). The camera is calibrated using a camera calibration board or an object of known size. By capturing images of the calibration board from multiple angles, the camera's intrinsic and extrinsic parameters are calculated using Zhang Zhengyou's calibration method. These parameters are recorded and input into the data processing and analysis system.

[0170] Using the camera's intrinsic and extrinsic parameters and the two-dimensional coordinates of feature points, the three-dimensional coordinates of the feature points are calculated through the inverse process of perspective projection. A perspective projection equation is established based on the camera's intrinsic and extrinsic parameters and the two-dimensional coordinates of the feature points. Then, the three-dimensional coordinates of the feature points are calculated by solving the inverse process of the perspective projection equation. During the calculation process, camera calibration errors and image processing errors need to be considered, and necessary corrections and filtering are performed on the calculated three-dimensional coordinates.

[0171] (6) Construction of 3D Channel: Connect the converted 3D coordinate points into lines or surfaces to form a 3D spatial discharge channel. Connect the converted 3D coordinate points into lines or surfaces in a certain order. Use 3D modeling software or programming languages ​​(such as Python) to construct the 3D channel. When connecting coordinate points, the shape and continuity of the discharge channel need to be considered to ensure that the constructed 3D channel conforms to the actual discharge situation.

[0172] Given that the vertical distance between the discharge start and end points is 10m, the actual length corresponding to each cell in the matrix can be calculated. Referring to the coordinates of the leader discharge start point, the coordinates of each point in the leader channel can be calculated. Combining this with the vertical distance of the leader over time obtained by the high-speed camera, the three-dimensional dynamic evolution path of the leader can be reconstructed, such as... Figure 6 As shown.

[0173] Combination Figure 5 The above three-dimensional reconstruction process is described below:

[0174] For the matching angle combinations (α1, β1) and (α2, β2), let L be the corresponding lightning channel point, and its three-dimensional coordinates be (L... X L Y L Z Easily obtained:

[0175]

[0176]

[0177] For optical observations, the greater the observation distance, the greater the observation error. Therefore, the final 3D reconstructed point L is located closer to the end of line segment L1L2 where the observation distance is shorter.

[0178]

[0179]

[0180]

[0181] P is the projection point of the 3D reconstructed point L onto the Z=0 plane.

[0182]

[0183]

[0184]

[0185] We can obtain:

[0186]

[0187]

[0188] Where L X L Y and L Z These are the 3D reconstructed coordinates of L. By reconstructing all matching angle combinations according to the above process, a set of 3D point coordinate sequences can be obtained. Connecting these sequences yields the reconstructed 3D lightning channel.

[0189] In one embodiment, step S6 can be implemented through the following process:

[0190] (1) Time synchronization and interpolation: Based on the shooting frame rate of the high-speed camera, the three-dimensional discharge channels at adjacent moments are time synchronized and interpolated to construct a smooth dynamic three-dimensional discharge channel.

[0191] (2) Results output and display: The constructed dynamic three-dimensional discharge channel will be output and displayed in the form of a three-dimensional model or animation to facilitate subsequent analysis and research.

[0192] Based on the same inventive concept, this application also provides a device for implementing the above-mentioned method for three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations in the embodiments of the device for three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras provided below can be found in the limitations of the method for three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras described above, and will not be repeated here.

[0193] Please see Figure 7 This invention provides a three-dimensional adaptive reconstruction device for dynamic discharge channels based on multiple high-speed cameras, comprising:

[0194] The shooting point location determination module is used to determine the shooting points of multiple high-speed cameras with the discharge gap as the center.

[0195] The image acquisition module is used to capture dynamic images of long air gap discharges at various shooting points using a high-speed camera.

[0196] The image recognition module is used to process the images captured at each shooting point, identify the discharge channels in the images, and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images;

[0197] The three-dimensional coordinate determination module is used to determine the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap;

[0198] The image matching and processing module is used to match the images of each discharge channel captured at different shooting points and at different times, and to construct the discharge channel in three-dimensional space at different times based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image.

[0199] The 3D reconstruction module is used to construct a complete dynamic 3D discharge channel based on the 3D discharge channel of each discharge channel at different times.

[0200] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0201] Reference Figure 8 The present invention also provides a computer device, including: a memory and a processor, and a computer program stored in the memory. When the computer program is executed on the processor, it implements the dynamic discharge channel three-dimensional adaptive reconstruction method based on multiple high-speed cameras as described in any of the above methods.

[0202] The computer device may be a desktop computer, laptop, handheld computer, or cloud server, etc. This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 8 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. They may include more or fewer components than shown in the illustration, or combinations of certain components, or different components. For example, they may also include input / output devices, network access devices, etc.

[0203] The processor referred to can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0204] In some embodiments, the memory may be an internal storage unit of the computer device, such as a hard drive or RAM. In other embodiments, the memory may be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory may include both internal and external storage units of the computer device. The memory is used to store the operating system, applications, bootloader, data, and other programs, such as the program code of the computer program. The memory can also be used to temporarily store data that has been output or will be output.

[0205] This invention also provides a computer-readable storage medium storing a computer program thereon. When the computer program is run by a processor, it implements the dynamic discharge channel three-dimensional adaptive reconstruction method based on multiple high-speed cameras as described in any of the above methods.

[0206] In this embodiment, if the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0207] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0208] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0209] In the embodiments disclosed in this application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0210] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for three-dimensional adaptive reconstruction of dynamic discharge channels based on multiple high-speed cameras, characterized in that, Includes the following steps: Using the discharge gap as the center, determine the shooting points of multiple high-speed cameras; At each of the aforementioned shooting points, dynamic images of long air gap discharge are captured using the high-speed camera. The images captured at each of the aforementioned shooting points are processed to identify the discharge channels in the images and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images; Determine the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap; The discharge channel images captured at different shooting points and at different times for each discharge channel are matched, and the discharge channel in three-dimensional space at different times is constructed based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image. A complete dynamic three-dimensional discharge channel is constructed based on the discharge channels in the three-dimensional space at different times for each discharge channel; Matching the discharge channel images captured at different shooting points and at different times for each of the aforementioned discharge channels includes: Based on the light intensity distribution characteristics of the discharge channel, the threshold for feature point detection is dynamically adjusted, and a local contrast analysis method is used to achieve adaptive recognition of regions with different light intensities. Using morphological filtering or edge detection algorithms, possible bright areas and edges on the discharge channel are initially screened as candidate feature points; Taking into account factors such as the stability, significance, and uniformity of distribution of feature points, high-quality feature points are selected from the candidate feature points. Nearest neighbor matching or random sampling consensus matching is used to match feature points in different frames; the matching rate of each feature point in different frames is calculated, and feature points with higher matching rates are selected as stable feature points. The saliency of feature points is evaluated based on their brightness and contrast; feature points with higher brightness and contrast are selected as saliency feature points. Spatial distribution statistics or cluster analysis are used to calculate the positional distribution of each feature point on the discharge channel; based on the positional distribution, the uniformity of the feature point distribution is evaluated; if the feature points in some areas are too dense, some redundant feature points are deleted; if the feature points in some areas are too sparse, some new feature points are added to ensure that the feature points are evenly distributed on the discharge channel. Based on the Kalman filter tracking method, and combined with the motion characteristics of the discharge channel, the selected feature points are stably tracked. During the tracking process, the feature point descriptors are updated in real time to adapt to the dynamic changes of the discharge channel.

2. The three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras according to claim 1, characterized in that, The images captured at each of the aforementioned shooting points are processed to identify discharge channels in the images and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images, including: The images captured at each of the aforementioned shooting points are converted into grayscale images; The grayscale image is processed using a grayscale image processing method based on morphological filtering to obtain the upper and lower envelope surfaces of the grayscale image; The upper and lower envelope surfaces are smoothed using a mean filter to obtain smoothed upper and lower envelope surfaces; The characteristics of the discharge channel are identified based on the smooth upper and lower envelope surfaces, and the grayscale image is processed into a binary image while retaining the characteristics. The discharge channel is identified in the binary image, and a two-dimensional coordinate sequence of the discharge channel in the image is constructed.

3. The three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras according to claim 2, characterized in that, The upper and lower envelope surfaces are determined according to the following formula: ； ； In the formula, and Pixel positions The upper and lower envelope surface values ​​at the location; In pixel position Centered The collection of pixels in a window; Original grayscale image At pixel position The gray value at the location; ⊕ is the morphological dilation filter; ⊖ is the morphological erosion filter; and b is the binary indicator function.

4. The three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras according to claim 3, characterized in that, The smooth upper and lower envelope surfaces are determined according to the following formula: ； ； In the formula, and Pixel positions The smoothed upper and lower envelope surface values; and Pixel positions The upper and lower envelope surface values ​​at the location.

5. The three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras according to claim 1, characterized in that, Matching the discharge channel images captured at different shooting points for each discharge channel, including: Extract a point from the path of the discharge channel as a reference point and obtain the two-dimensional coordinates of the reference point in each discharge channel image; the two-dimensional coordinates include horizontal coordinates and vertical coordinates; Based on the three-dimensional coordinates of the shooting point corresponding to the discharge channel image and the two-dimensional coordinates of the reference point, calculate the azimuth and elevation angles of the reference point observed from the shooting point. In the same three-dimensional coordinate system, any two shooting points are observed along the perpendicular line to the common reference point according to the corresponding azimuth and elevation angles, to obtain two observation points on the perpendicular line of the two shooting points; the common reference point is the point determined by projecting the two horizontal coordinates of the reference point onto the horizontal plane in the discharge channel image taken by the two shooting points. The distance difference between the two observation points is used to determine whether the images of the discharge channel captured by the two shooting points match. If they match, they are assigned to the image set of the corresponding discharge channel. If not, the shooting point is changed until all shooting points have been traversed.

6. The three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras according to claim 5, characterized in that, When determining whether the discharge channel images captured by the two shooting points match based on the distance difference between the two observation points, the distance difference between the two observation points is converted into an elevation angle error before the judgment is made. The elevation angle error is calculated according to the following formula: ； In the formula, The elevation angle error; The distance difference; First shooting point Distance from the common reference point Perpendicular line Horizontal distance, First shooting point The horizontal plane and the vertical line The intersection; Second shooting location Distance from the common reference point Perpendicular line Horizontal distance, Second shooting location The horizontal plane and the vertical line The intersection; The distance difference is determined according to the following formula: ； ； ； ； ； In the formula, and The first shooting point is respectively According to the corresponding azimuth and elevation angles and the second shooting point According to the corresponding azimuth and elevation angles on the vertical line The two observation points observed above; and These are the common reference points. Distance between two observation points and The distance; and These are the distances between the two intersection points and the two observation points, respectively; and The first shooting point is respectively and the second shooting point Height; and These are the distances between the points projected from the two shooting points onto the horizontal plane where the common reference point is located, and the common reference point itself.

7. The three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras according to claim 6, characterized in that, Based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image, discharge channels in three-dimensional space at different times are constructed, including: For any reference point in the discharge channel Using the three-dimensional coordinates of at least two shooting points matched by the reference point, the distances from the two shooting points to the corresponding two observation points are calculated. and ; Based on distance and Calculate the reference point The distances to the two observation points are as follows: ； ； In the formula, and The reference points are respectively Distance to the two observation points; For the reference point Its three-dimensional reconstruction coordinates are determined according to the following formula: ； ； ； In the formula, , and The coordinates are the three-dimensional reconstruction coordinates; and The reference point is determined from the discharge channel images captured from the common reference point to the first shooting point and the second shooting point, respectively. Horizontal coordinates; The three-dimensional reconstructed coordinates of all reference points are combined to form the three-dimensional reconstructed coordinates of the discharge channel.

8. A three-dimensional adaptive reconstruction device for dynamic discharge channels based on multiple high-speed cameras, characterized in that, include: The shooting point location determination module is used to determine the shooting points of multiple high-speed cameras with the discharge gap as the center. The image acquisition module is used to capture dynamic images of long air gap discharge at each of the aforementioned shooting points using the high-speed camera. The image recognition module is used to process the images captured at each of the shooting points, identify the discharge channels in the images, and construct a two-dimensional coordinate sequence of the discharge channels in the corresponding images; A three-dimensional coordinate determination module is used to determine the three-dimensional coordinate position of each shooting point in a three-dimensional spatial coordinate system centered on the discharge gap; The image matching and processing module is used to match the discharge channel images captured at different shooting points and at different times for each discharge channel, and to construct the discharge channel in three-dimensional space at different times based on the two-dimensional coordinate sequence of each discharge channel in the matched image and the three-dimensional coordinate position of the shooting point corresponding to the matched image. The three-dimensional reconstruction module is used to construct a complete dynamic three-dimensional discharge channel based on the discharge channels in the three-dimensional space at different times for each discharge channel; Matching the discharge channel images captured at different shooting points and at different times for each of the aforementioned discharge channels includes: Based on the light intensity distribution characteristics of the discharge channel, the threshold for feature point detection is dynamically adjusted, and a local contrast analysis method is used to achieve adaptive recognition of regions with different light intensities. Using morphological filtering or edge detection algorithms, possible bright areas and edges on the discharge channel are initially screened as candidate feature points; Taking into account factors such as the stability, significance, and uniformity of distribution of feature points, high-quality feature points are selected from the candidate feature points. Nearest neighbor matching or random sampling consensus matching is used to match feature points in different frames; the matching rate of each feature point in different frames is calculated, and feature points with higher matching rates are selected as stable feature points. The saliency of feature points is evaluated based on their brightness and contrast; feature points with higher brightness and contrast are selected as saliency feature points. Spatial distribution statistics or cluster analysis are used to calculate the positional distribution of each feature point on the discharge channel; based on the positional distribution, the uniformity of the feature point distribution is evaluated; if the feature points in some areas are too dense, some redundant feature points are deleted; if the feature points in some areas are too sparse, some new feature points are added to ensure that the feature points are evenly distributed on the discharge channel. Based on the Kalman filter tracking method, and combined with the motion characteristics of the discharge channel, the selected feature points are stably tracked. During the tracking process, the feature point descriptors are updated in real time to adapt to the dynamic changes of the discharge channel.

9. A computer device, characterized in that, The device includes a processor and a memory: The memory is used to store computer programs and send the instructions of the computer programs to the processor; The processor executes, according to the instructions of the computer program, a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements a three-dimensional adaptive reconstruction method for dynamic discharge channels based on multiple high-speed cameras as described in any one of claims 1-7.

Images