A method, system, device and storage medium for detecting defects in power cables

Abstract

The present invention discloses a method, system, device and storage medium for power cable defect detection, the method comprising: collecting power cable samples; inputting a detection signal at the head end of the power cable sample, collecting a feedback signal at the tail end of the power cable sample, and generating a preliminary defect detection result; collecting the power cable sample image based on a multi-view camera and performing preprocessing; performing two-dimensional defect detection to obtain a two-dimensional defect detection result of the power cable sample; reconstructing a three-dimensional model of the power cable sample, and performing three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample; fusing and analyzing the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate a final defect detection result. The present invention improves the accuracy and efficiency of power cable defect detection, reduces labor costs, and thereby improves the reliability of power transportation based on power cables.

Description

Technical Field

[0001] The present invention relates to the technical field of power cable detection, and in particular to a method, system, device and storage medium for power cable defect detection. Background Art

[0002] With the rapid development of urbanization and economic level in my country, the power system is becoming more and more important in life, and the mode of power transmission through power cables has been widely used. When transporting electricity over long distances, in order to minimize losses, it is necessary to perform a boost operation before transmitting through power cables. The length of a single section of power cable is limited, so an intermediate head is required for series connection. Whether it is for power cables erected in the sky or laid underground, due to the high voltage transmitted, they are in a state of heavy overload and overheating for a long time, and the working environment of the power cable is either in the air or underground, and it is easy to cause damage under the sun and rain, which in turn affects the insulation performance, resulting in large losses in the power cable or even failure to supply power.

[0003] When a fault occurs, it is necessary to perform defect detection on the faulty power cable to determine the cause of the fault. The existing commonly used defect detection method is based on manual judgment. Power maintenance personnel look for faults through visual observation. The labor cost is high, and it is affected by factors such as the power maintenance personnel's own attention, fatigue, and even professionalism. It has a certain degree of uncontrollability, and some defects cannot be found through naked eye observation by power maintenance personnel. Therefore, the existing defect detection method is not mature, and there is no perfect defect detection standard, so it is impossible to obtain accurate defect judgment, and the efficiency of defect detection is low. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art. The present invention provides a method, system, device and storage medium for power cable defect detection, which improves the accuracy and efficiency of power cable defect detection, reduces labor costs, and thus improves the reliability of power transportation based on power cables.

[0005] The present invention provides a method for detecting defects in a power cable, the method comprising:

[0006] Collecting power cable samples suspected of having defects based on a power cable defect detection request;

[0007] Inputting a detection signal at the head end of the power cable sample, collecting a feedback signal at the tail end of the power cable sample, and generating a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal;

[0008] Collecting the power cable sample images based on a multi-view camera, and preprocessing the power cable sample images to obtain the preprocessed power cable sample images;

[0009] Performing two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample;

[0010] Reconstructing a three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and performing three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample;

[0011] The two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample are fused and analyzed to generate the final defect detection result of the power cable sample.

[0012] Further, generating a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal includes:

[0013] Establishing an autocorrelation function of the detection signal and the feedback signal;

[0014] Calculate the time difference between inputting the detection signal at the head end of the power cable sample and collecting the feedback signal at the tail end of the power cable sample;

[0015] Substituting the time difference into the autocorrelation function of the detection signal and the feedback signal to generate a cross-correlation function of the detection signal and the feedback signal;

[0016] Establishing a nonlinear resonance equation of the detection signal and the feedback signal based on the cross-correlation function of the detection signal and the feedback signal, and calculating the correlation amplitude of the detection signal and the feedback signal;

[0017] A preliminary defect detection result of the power cable sample is generated based on the correlation amplitude of the detection signal and the feedback signal.

[0018] Furthermore, the generating of the preliminary defect detection result of the power cable sample based on the correlation amplitude of the detection signal and the feedback signal includes:

[0019] When the correlation amplitude between the detection signal and the feedback signal is zero, it is determined that the power cable sample has no defects;

[0020] When the correlation amplitude between the detection signal and the feedback signal is not zero, it is determined that the power cable sample has a defect.

[0021] Furthermore, the collecting the power cable sample images based on the multi-view camera and preprocessing the power cable sample images to obtain the preprocessed power cable sample images includes:

[0022] Setting a multi-view camera, and collecting multi-view images of the power cable sample based on the multi-view camera;

[0023] The collected multi-view images of the power cable samples are filtered, geometrically corrected, image fused, and sharpened to obtain the preprocessed multi-view images of the power cable samples.

[0024] Furthermore, performing two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample includes:

[0025] After edge extraction is performed on the preprocessed power cable sample image, feature information is extracted based on a deep neural network, and feature analysis and regression classification are performed to obtain a two-dimensional defect detection result of the power cable sample.

[0026] Furthermore, reconstructing the three-dimensional model of the power cable sample based on the preprocessed power cable sample image and performing three-dimensional defect detection to obtain the three-dimensional defect detection result of the power cable sample includes:

[0027] Performing camera-laser calibration based on the position information of the multi-view camera;

[0028] Extracting point cloud information of the preprocessed power cable sample image, and performing point cloud supervoxel processing to obtain simplified point cloud information;

[0029] Determine the voxel center of the point cloud and calculate the curvature of the point cloud;

[0030] Defects of the three-dimensional model of the power cable sample are detected based on the point cloud curvature.

[0031] Further, the camera-laser calibration based on the position information of the multi-view camera includes:

[0032] Establish camera-laser calibration coordinate system;

[0033] Performing binocular positioning on any two cameras of the multi-view cameras respectively to obtain a plurality of binocular positioning information;

[0034] Normalizing the plurality of binocular positioning information to obtain multi-target positioning information;

[0035] Extracting the laser center line of the multi-view camera and calibrating the laser plane to obtain laser calibration information;

[0036] The multi-target calibration information and the laser calibration information are integrated to obtain camera-laser calibration information.

[0037] The present invention also provides a system for detecting defects in a power cable, the system for detecting defects in a power cable is used to implement the above-mentioned method for detecting defects in a power cable, and the system for detecting defects in a power cable comprises:

[0038] A sample collection module, the sample collection module is used to collect samples of power cables suspected of having defects based on a power cable defect detection request;

[0039] A preliminary defect detection module, the preliminary defect detection module is used to input a detection signal at the head end of the power cable sample, collect a feedback signal at the tail end of the power cable sample, and generate a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal;

[0040] An image acquisition processing module, the image acquisition processing module is used to collect the power cable sample images based on a multi-view camera, and pre-process the power cable sample images to obtain the pre-processed power cable sample images;

[0041] A two-dimensional defect detection module, wherein the two-dimensional defect detection module is used to perform two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample;

[0042] A three-dimensional defect detection module, the three-dimensional defect detection module is used to reconstruct a three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and perform three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample;

[0043] A defect detection fusion module is used to fuse and analyze the two-dimensional defect detection results and the three-dimensional defect detection results of the power cable sample to generate a final defect detection result of the power cable sample.

[0044] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned method for detecting defects in power cables.

[0045] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned method for detecting defects in a power cable is implemented.

[0046] The present invention provides a method, system, device and storage medium for power cable defect detection, which collects power cable samples suspected of having defects, inputs detection signals at the head end and collects feedback signals at the tail end, and determines whether defects exist through preliminary analysis, which is beneficial to improving the efficiency of subsequent defect detection; based on the collection of images by a multi-view camera, an all-round image of the power cable sample can be obtained, and the obtained image is preprocessed to improve the accuracy of defect detection; two-dimensional and three-dimensional defect detection are performed on the power cable samples respectively, which improves the accuracy and efficiency of power cable defect detection, reduces labor costs, and thereby improves the reliability of power transportation based on power cables. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 is a flow chart of a method for detecting power cable defects in Embodiment 1 of the present invention;

[0049] Figure 2 is a flow chart of generating preliminary defect detection results of a power cable sample in the first embodiment of the present invention;

[0050] Figure 3 is a flow chart of collecting and preprocessing power cable sample images in the first embodiment of the present invention;

[0051] Figure 4 is a flow chart of performing three-dimensional defect detection on a power cable sample in Embodiment 1 of the present invention;

[0052] Figure 5 is a flowchart of performing camera-laser calibration in Embodiment 1 of the present invention;

[0053] Figure 6 It is a system architecture diagram of power cable defect detection in the second embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] In the present invention, it should be understood that terms such as "include" or "have" are intended to indicate the existence of features, numbers, steps, behaviors, components, parts or a combination thereof disclosed in the specification, and are not intended to exclude the possibility that one or more other features, numbers, steps, behaviors, components, parts or a combination thereof exist or are added.

[0056] It should also be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0057] Embodiment 1

[0058] Embodiment 1 of the present invention provides a method for detecting defects in a power cable, the method comprising: collecting a sample of a power cable suspected of having defects based on a power cable defect detection request; inputting a detection signal at the head end of the power cable sample, collecting a feedback signal at the tail end of the power cable sample, and generating a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal; collecting an image of the power cable sample based on a multi-view camera, and preprocessing the image of the power cable sample to obtain the preprocessed power cable sample image; performing two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample; reconstructing a three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and performing three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample; fusing and analyzing the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate a final defect detection result of the power cable sample.

[0059] In an optional implementation of this embodiment, Figure 1 As shown, Figure 1 The flowchart of the method for detecting power cable defects in the first embodiment of the present invention is shown, comprising the following steps:

[0060] S101, collecting power cable samples suspected of having defects based on a power cable defect detection request;

[0061] In an optional implementation of the present embodiment, abnormal operating data of the power cable is collected, analyzed, and compared with the normal operating data during normal operation to determine whether there is an abnormality. If it is determined that there is an abnormality, a power cable defect detection request is generated, and power cable samples suspected of defects are collected based on the power cable defect detection request.

[0062] In an optional implementation of this embodiment, samples of power cables suspected of having defects are collected manually, by robots, or by drones.

[0063] S102, inputting a detection signal at the head end of the power cable sample, collecting a feedback signal at the tail end of the power cable sample, and generating a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal;

[0064] In an optional implementation of the present embodiment, after collecting a power cable sample suspected of having defects, one end is calibrated as the head end and the other end is calibrated as the tail end, a detection signal is generated, and the detection signal is input at the head end of the power cable sample, and a feedback signal is collected at the tail end of the power cable sample.

[0065] In an optional implementation of this embodiment, the detection signal is a stepped frequency signal in a sinusoidal signal.

[0066] In an optional implementation of this embodiment, after the feedback signal is collected, the detection signal and the feedback signal are analyzed to generate a preliminary defect detection result of the power cable sample.

[0067] Specifically, Figure 2 As shown, Figure 2 The flowchart of generating preliminary defect detection results of a power cable sample in the first embodiment of the present invention is shown, and includes the following steps:

[0068] S201, establishing an autocorrelation function of the detection signal and the feedback signal;

[0069] In an optional implementation of this embodiment, the autocorrelation function of the detection signal and the feedback signal is expressed as follows:

[0070]

[0071] In the formula, is the autocorrelation function of the detection signal and the feedback signal, D(t) is the detection signal, and F(t) is the feedback signal.

[0072] S202, calculating the time difference between inputting the detection signal at the head end of the power cable sample and collecting the feedback signal at the tail end of the power cable sample;

[0073] In an optional implementation of this embodiment, since the detection signal and the feedback signal are essentially the same, there is only a certain difference in input time and collection time. The expression for the time difference between inputting the detection signal at the head end of the power cable sample and collecting the feedback signal at the tail end of the power cable sample is as follows:

[0074]

[0075] Where, t′ is the time difference, r is the reflectivity of the feedback signal, S(t) is the incident amplitude of the detection signal, v is the signal transmission speed, α(t) is the round-trip time from the beginning to the end of the power cable sample, is the power ratio of the detection signal to the feedback signal.

[0076] S203, substituting the time difference into the autocorrelation function of the detection signal and the feedback signal to generate a cross-correlation function of the detection signal and the feedback signal;

[0077] In an optional implementation of this embodiment, the cross-correlation function of the detection signal and the feedback signal is expressed as follows:

[0078]

[0079] In the formula, is the cross-correlation function of the detection signal and the feedback signal, D(t) is the detection signal, F(t) is the feedback signal, S(t) is the incident amplitude of the detection signal, α(t) is the round-trip time from the head end to the tail end of the power cable sample, is the power ratio of the detection signal to the feedback signal.

[0080] S204, establishing a nonlinear resonance equation of the detection signal and the feedback signal based on the cross-correlation function of the detection signal and the feedback signal, and calculating the correlation amplitude of the detection signal and the feedback signal;

[0081] In an optional implementation of this embodiment, the expression of the nonlinear resonance equation of the detection signal and the feedback signal is as follows:

[0082]

[0083] Where d is the correlation amplitude between the detection signal and the feedback signal, α(t) is the round-trip time from the beginning to the end of the power cable sample, is the power ratio of the detection signal to the feedback signal, D(t) is the detection signal, F(t) is the feedback signal, and δ is the damping coefficient, that is, the damping ratio of the detection signal to the feedback signal.

[0084] S205: Generate a preliminary defect detection result of the power cable sample based on the correlation amplitude between the detection signal and the feedback signal.

[0085] In an optional implementation of this embodiment, when the correlation amplitude between the detection signal and the feedback signal is zero, it is judged that the power cable sample has no defects; when the correlation amplitude between the detection signal and the feedback signal is not zero, it is judged that the power cable sample has defects.

[0086] Specifically, when the correlation amplitude d between the detection signal and the feedback signal is 0, it is determined that the power cable sample has no defects.

[0087] Furthermore, when the correlation amplitude d≠0 between the detection signal and the feedback signal, it is determined that the power cable sample has a defect.

[0088] S103, collecting the power cable sample images based on a multi-view camera, and preprocessing the power cable sample images to obtain preprocessed power cable sample images;

[0089] In an optional implementation of the present embodiment, when it is determined that the power cable sample has defects in the preliminary defect detection result of the power cable sample generated in step S102, the power cable sample image is collected based on a multi-view camera and the power cable sample image is preprocessed.

[0090] In an optional implementation of this embodiment, Figure 3 As shown, Figure 3 The flowchart of collecting and preprocessing power cable sample images in the first embodiment of the present invention is shown, including the following steps:

[0091] S301, setting a multi-view camera, and collecting multi-view images of the power cable sample based on the multi-view camera;

[0092] In an optional implementation of this embodiment, a multi-view camera including at least three mutually perpendicular directions (including x, y, z) is provided, and at least three mutually perpendicular multi-view images of the power cable sample are collected based on the multi-view camera.

[0093] S302, filtering the multi-view images of the collected power cable samples;

[0094] In an optional implementation of this embodiment, a median filtering method is used to filter the multi-view images of the collected power cable samples, and the expression is as follows:

[0095] b(x,y,z)=med(a(xi,yj,zk));

[0096] Where b(x, y, z) is the pixel value of the image after filtering, a(x, y, z) is the pixel value of the image before filtering, and i, j, k are the correction values ​​of the filter coefficients.

[0097] S303, performing geometric correction on the multi-view images of the power cable sample after filtering;

[0098] In an optional implementation of this embodiment, according to the difference between the orientations and angles of the set multi-view cameras, the geometric error is calculated in combination with the multi-view camera parameters, and the multi-view image of the filtered power cable sample is geometrically corrected based on the geometric error.

[0099] S304, performing image fusion on the multi-view images of the power cable sample after geometric correction processing;

[0100] In an optional implementation of this embodiment, it is determined whether there are overlapping areas in the multi-view images of the power cable sample after the geometric correction process, and the overlapping areas are extracted and merged at interval points.

[0101] S305, sharpening the multi-view images of the power cable samples after the image fusion processing.

[0102] In an optional implementation of this embodiment, the multi-view images of the power cable sample after image fusion processing based on the Laplacian image sharpening operator are sharpened, and the expression is as follows:

[0103] g(x,y,z)=∑∑f(xr,ys,zt)H(r,s,t);

[0104] Where H(r,s,t) is the Laplace template operator, f(x,y,z) is the pixel value of the image before sharpening, and g(x,y,z) is the pixel value of the image after sharpening.

[0105] S104, performing two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample;

[0106] In an optional implementation of the present embodiment, after edge extraction is performed on the preprocessed power cable sample image, feature information is extracted based on a deep neural network, and feature analysis and regression classification are performed to obtain a two-dimensional defect detection result of the power cable sample.

[0107] In an optional implementation of this embodiment, edge extraction is performed on the preprocessed power cable sample image based on a canny edge extraction algorithm.

[0108] In an optional implementation of this embodiment, feature information is extracted based on the YOLOv5 deep neural network, and feature analysis and regression classification are performed.

[0109] Specifically, shallow features of the power cable sample image after edge extraction are extracted based on the convolution layer, and the dimension of the shallow features is increased based on the pooling layer to obtain deep features.

[0110] Furthermore, the deep features are analyzed based on the fully connected layer, different labels are calibrated, and the deep features calibrated with the same label are regressed and classified into a class of feature vectors.

[0111] It should be noted that the gaps that occur during regression classification are filled based on the activation function, and the error between the predicted value and the actual value is calculated based on the loss function.

[0112] In an optional implementation of this embodiment, after the final feature vector is obtained, a feature vector graph is drawn based on the feature vector, and a two-dimensional defect detection result of the power cable sample is obtained based on the feature vector graph.

[0113] S105, reconstructing a three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and performing three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample;

[0114] In an optional implementation of this embodiment, Figure 4 As shown, Figure 4 A flow chart of performing three-dimensional defect detection on a power cable sample in Embodiment 1 of the present invention is shown, comprising the following steps:

[0115] S401, performing camera-laser calibration based on position information of the multi-view camera;

[0116] In an optional implementation of this embodiment, Figure 5 As shown, Figure 5 A flowchart of performing camera-laser calibration in Embodiment 1 of the present invention is shown, comprising the following steps:

[0117] S501, establishing a camera-laser calibration coordinate system;

[0118] In an optional implementation of this embodiment, the world coordinate system is determined as the reference coordinate system, the reference coordinates of the position information of the multi-view camera in the world coordinate system are obtained, and then the camera coordinate system of each camera is established, as well as the conversion relationship with the world coordinate system.

[0119] In an optional implementation of this embodiment, the laser plane and the transformation relationship between the laser plane and the world coordinate system are determined to generate a laser coordinate system.

[0120] S502, performing binocular positioning on any two cameras in the multi-view cameras respectively to obtain a number of binocular positioning information;

[0121] In an optional implementation of this embodiment, a multi-view camera with three axes of x, y, and z is used.

[0122] In an optional implementation of this embodiment, firstly, the x-axis and y-axis cameras are binocularly positioned, and the projection relationship of the two cameras under binocular positioning is:

[0123]

[0124] In the formula, Q 1 is the rotation and translation matrix of the camera on the x-axis, Q 2 The rotation and translation matrix of the camera on the y-axis.

[0125] Combining the above formulas, the conversion formula between the camera coordinate system of the multi-view camera located on the y-axis and the camera coordinate system of the multi-view camera located on the x-axis is obtained as follows:

[0126] Q 21 =Q 1 Q 2 -1 ;

[0127] In the formula, Q 11 is the conversion relationship between the camera coordinate system of the multi-view camera located on the y-axis and the camera coordinate system of the multi-view camera located on the x-axis.

[0128] The same calculation process as above, the conversion formula between the camera coordinate system of the multi-view camera located on the z-axis and the camera coordinate system of the multi-view camera located on the x-axis is obtained as follows:

[0129] Q 31 =Q 1 Q 3 -1 ;

[0130] In the formula, Q 31 It is the conversion relationship between the camera coordinate system of the multi-view camera located on the z-axis and the camera coordinate system of the multi-view camera located on the x-axis.

[0131] S503, normalizing the plurality of dual-target positioning information to obtain multi-target positioning information;

[0132] In an optional implementation of the present embodiment, after calculating the mutual conversion relationship of the camera coordinate systems of the multi-view cameras located on the x-axis, y-axis and z-axis, they are normalized to the global camera coordinate system, and multiple iterative optimizations are performed based on the LM estimation method to reduce the multi-target positioning error, and then the multi-target positioning information is generated.

[0133] S504, extracting the laser center line of the multi-view camera, and calibrating the laser plane to obtain laser calibration information;

[0134] In an optional implementation of this embodiment, Gaussian coefficients are set, multivariate recursive coefficients generated based on the Gaussian coefficients are calculated, a Hessian matrix is ​​constructed, eigenvalues ​​and eigenvectors in the Hessian matrix are extracted, and the laser coordinate system of the laser center line of the multi-view camera is obtained.

[0135] In an optional implementation of this embodiment, the laser plane is determined, the laser plane is calibrated based on the laser coordinate system, and laser calibration information is acquired.

[0136] S505 , integrating the multi-target calibration information and the laser calibration information to obtain camera-laser calibration information.

[0137] In an optional implementation of this embodiment, the conversion relationship between the laser coordinate system and the global camera coordinate system is determined, and the calibration error is calculated to obtain camera-laser calibration information.

[0138] S402, extracting point cloud information of the preprocessed power cable sample image, and performing point cloud supervoxelization processing to obtain simplified point cloud information;

[0139] In an optional implementation of this embodiment, the seed voxel is determined, the point cloud information of the preprocessed power cable sample image is extracted, the characteristic distance metric between each point cloud voxel information is calculated, and the constrained flow clustering algorithm is used to perform clustering calculation to obtain simplified point cloud information.

[0140] Specifically, the expression of the feature distance metric is as follows:

[0141]

[0142] In the formula, D is the feature distance metric, γ 1 is the point cloud spatial position influencing factor, D 1 is the spatial position similarity of the point cloud, γ 2 is the point cloud space curvature influence factor, D 2 is the point cloud space bending similarity, γ 2 is the point cloud feature influencing factor, D 3 is the point cloud feature similarity, R is the position resolution, and C is the bending resolution.

[0143] S403, determining the voxel center of the point cloud and calculating the curvature of the point cloud;

[0144] In an optional implementation of this embodiment, the simplified point cloud information is analyzed, the point cloud voxel center is determined, and the point cloud curvature is calculated. The expression is as follows:

[0145]

[0146] Where k(T) is the point cloud curvature, N is the normal vector of the point cloud voxel center coordinate, and p i is the coordinate of any point in the direction of the normal vector of the point cloud voxel center, p 0 is the center coordinate of the point cloud voxel.

[0147] S404: Detect defects of the three-dimensional model of the power cable sample based on the point cloud curvature.

[0148] In an optional implementation of this embodiment, a point cloud curvature threshold is set, and when the point cloud curvature exceeds the point cloud curvature threshold, it is determined that a three-dimensional defect exists.

[0149] S106: Fusing and analyzing the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate a final defect detection result of the power cable sample.

[0150] In an optional implementation of this embodiment, after aligning the two-dimensional defect coordinate system in the two-dimensional defect detection result with the three-dimensional defect coordinate system in the three-dimensional defect detection result, a projection relationship expression is generated, and a fusion analysis is performed based on the projection relationship expression.

[0151] It should be noted that, in general, there are more two-dimensional defects in the two-dimensional defect detection results than three-dimensional defects in the three-dimensional defect detection results. This is because the defect depth is not considered in the two-dimensional defect detection results. Therefore, when both two-dimensional defects and three-dimensional defects exist at a certain defect location, it is determined that there is a defect at that location.

[0152] In summary, embodiment 1 of the present invention provides a method for detecting defects in power cables, which collects power cable samples suspected of having defects, inputs detection signals at the head end and collects feedback signals at the tail end, and determines whether defects exist through preliminary analysis, which is beneficial to improving the efficiency of subsequent defect detection; by collecting images based on a multi-view camera, a full-range image of the power cable sample can be obtained, and the obtained image is preprocessed to improve the accuracy of defect detection; two-dimensional and three-dimensional defect detection are performed on the power cable samples respectively, which improves the accuracy and efficiency of power cable defect detection, reduces labor costs, and thereby improves the reliability of power transportation based on power cables.

[0153] Embodiment 2

[0154] A second embodiment of the present invention provides a system for detecting defects in a power cable. The system for detecting defects in a power cable is used to implement the method for detecting defects in a power cable in the first embodiment.

[0155] In an optional implementation of this embodiment, Figure 6 As shown, Figure 6The system architecture diagram of power cable defect detection in the second embodiment of the present invention is shown, and the power cable defect detection system includes:

[0156] A sample collection module 10, the sample collection module 10 is used to collect samples of power cables suspected of having defects based on a power cable defect detection request;

[0157] A preliminary defect detection module 20, the preliminary defect detection module 20 is used to input a detection signal at the head end of the power cable sample, collect a feedback signal at the tail end of the power cable sample, and generate a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal;

[0158] An image acquisition processing module 30, wherein the image acquisition processing module 30 is used to collect the power cable sample images based on a multi-view camera, and pre-process the power cable sample images to obtain the pre-processed power cable sample images;

[0159] A two-dimensional defect detection module 40, the two-dimensional defect detection module 40 is used to perform two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample;

[0160] A three-dimensional defect detection module 50, the three-dimensional defect detection module 50 is used to reconstruct the three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and perform three-dimensional defect detection to obtain the three-dimensional defect detection result of the power cable sample;

[0161] The defect detection fusion module 60 is used to fuse and analyze the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate the final defect detection result of the power cable sample.

[0162] In summary, embodiment 2 of the present invention provides a system for power cable defect detection, which is used to implement the power cable defect detection method in embodiment 1, collect power cable samples suspected of having defects, input detection signals at the head end and collect feedback signals at the tail end, and determine whether there are defects through preliminary analysis, which is conducive to improving the subsequent defect detection efficiency; based on the collection of images by a multi-angle camera, a full-range image of the power cable sample can be obtained, and the obtained image is preprocessed to improve the accuracy of defect detection; two-dimensional and three-dimensional defect detection are performed on the power cable samples respectively, which improves the accuracy and efficiency of power cable defect detection, reduces labor costs, and thereby improves the reliability of power transportation based on power cables.

[0163] Embodiment 3

[0164] Embodiment 3 of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for detecting power cable defects in embodiment 1.

[0165] Embodiment 4

[0166] A fourth embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for detecting defects in a power cable in the first embodiment is implemented.

[0167] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0168] In addition, the embodiments of the present invention are described in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for detecting defects in a power cable, characterized in that: The method for detecting defects in a power cable comprises: Collecting power cable samples suspected of having defects based on a power cable defect detection request; Inputting a detection signal at the head end of the power cable sample, collecting a feedback signal at the tail end of the power cable sample, and generating a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal; The method of generating the preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal includes: establishing the autocorrelation function of the detection signal and the feedback signal; calculating the time difference between inputting the detection signal at the head end of the power cable sample and collecting the feedback signal at the tail end of the power cable sample; substituting the time difference into the autocorrelation function of the detection signal and the feedback signal to generate the cross-correlation function of the detection signal and the feedback signal; establishing the nonlinear resonance equation of the detection signal and the feedback signal based on the cross-correlation function of the detection signal and the feedback signal, and calculating the correlation amplitude of the detection signal and the feedback signal; generating the preliminary defect detection result of the power cable sample based on the correlation amplitude of the detection signal and the feedback signal; The generating of the preliminary defect detection result of the power cable sample based on the correlation amplitude of the detection signal and the feedback signal includes: when the correlation amplitude of the detection signal and the feedback signal is zero, judging that the power cable sample has no defect; when the correlation amplitude of the detection signal and the feedback signal is not zero, judging that the power cable sample has a defect; Collecting the power cable sample images based on a multi-view camera, and preprocessing the power cable sample images to obtain the preprocessed power cable sample images; Performing two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample; Reconstructing a three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and performing three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample; Performing a fusion analysis on the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate a final defect detection result of the power cable sample; The step of fusing and analyzing the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate the final defect detection result of the power cable sample comprises: aligning the two-dimensional defect coordinate system in the two-dimensional defect detection result with the three-dimensional defect coordinate system in the three-dimensional defect detection result, generating a projection relationship expression, and performing fusing and analysis based on the projection relationship expression; After aligning the two-dimensional defect coordinate system in the two-dimensional defect detection result with the three-dimensional defect coordinate system in the three-dimensional defect detection result, generating a projection relationship expression, and performing fusion analysis based on the projection relationship expression, including: when a two-dimensional defect and a three-dimensional defect exist at the same coordinate position in the projection relationship expression after the two-dimensional defect coordinate system is aligned with the three-dimensional defect coordinate system, it is determined that there is a defect at this coordinate position.

2. The method for detecting defects in a power cable according to claim 1, characterized in that: The collecting the power cable sample images based on the multi-view camera and preprocessing the power cable sample images to obtain the preprocessed power cable sample images includes: Setting a multi-view camera, and collecting multi-view images of the power cable sample based on the multi-view camera; The collected multi-view images of the power cable samples are filtered, geometrically corrected, image fused, and sharpened to obtain the preprocessed multi-view images of the power cable samples.

3. The method for detecting defects in a power cable according to claim 1, characterized in that: The performing two-dimensional defect detection on the preprocessed power cable sample image to obtain the two-dimensional defect detection result of the power cable sample includes: After edge extraction is performed on the preprocessed power cable sample image, feature information is extracted based on a deep neural network, and feature analysis and regression classification are performed to obtain a two-dimensional defect detection result of the power cable sample.

4. The method for detecting defects in a power cable according to claim 1, characterized in that: The reconstructing the three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and performing three-dimensional defect detection, and obtaining the three-dimensional defect detection result of the power cable sample includes: Performing camera-laser calibration based on the position information of the multi-view camera; Extracting point cloud information of the preprocessed power cable sample image, and performing point cloud supervoxel processing to obtain simplified point cloud information; Analyze the simplified point cloud information, determine the point cloud voxel center, and calculate the point cloud curvature; Defects of the three-dimensional model of the power cable sample are detected based on the point cloud curvature.

5. The method for detecting defects in a power cable according to claim 4, characterized in that: The camera-laser calibration based on the position information of the multi-view camera includes: Establish camera-laser calibration coordinate system; Performing binocular positioning on any two cameras of the multi-view cameras respectively to obtain a plurality of binocular positioning information; Normalizing the plurality of binocular positioning information to obtain multi-target positioning information; Extracting the laser center line of the multi-view camera and calibrating the laser plane to obtain laser calibration information; The multi-target calibration information and the laser calibration information are integrated to obtain camera-laser calibration information.

6. A system for detecting defects in power cables, characterized in that: The power cable defect detection system is used to implement the power cable defect detection method according to any one of claims 1 to 5, and the power cable defect detection system includes: A sample collection module, the sample collection module is used to collect samples of power cables suspected of having defects based on a power cable defect detection request; A preliminary defect detection module, the preliminary defect detection module is used to input a detection signal at the head end of the power cable sample, collect a feedback signal at the tail end of the power cable sample, and generate a preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal; The method of generating the preliminary defect detection result of the power cable sample based on the detection signal and the feedback signal includes: establishing the autocorrelation function of the detection signal and the feedback signal; calculating the time difference between inputting the detection signal at the head end of the power cable sample and collecting the feedback signal at the tail end of the power cable sample; substituting the time difference into the autocorrelation function of the detection signal and the feedback signal to generate the cross-correlation function of the detection signal and the feedback signal; establishing the nonlinear resonance equation of the detection signal and the feedback signal based on the cross-correlation function of the detection signal and the feedback signal, and calculating the correlation amplitude of the detection signal and the feedback signal; generating the preliminary defect detection result of the power cable sample based on the correlation amplitude of the detection signal and the feedback signal; The generating of the preliminary defect detection result of the power cable sample based on the correlation amplitude of the detection signal and the feedback signal includes: when the correlation amplitude of the detection signal and the feedback signal is zero, judging that the power cable sample has no defect; when the correlation amplitude of the detection signal and the feedback signal is not zero, judging that the power cable sample has a defect; An image acquisition processing module, the image acquisition processing module is used to collect the power cable sample images based on a multi-view camera, and pre-process the power cable sample images to obtain the pre-processed power cable sample images; A two-dimensional defect detection module, wherein the two-dimensional defect detection module is used to perform two-dimensional defect detection on the preprocessed power cable sample image to obtain a two-dimensional defect detection result of the power cable sample; A three-dimensional defect detection module, the three-dimensional defect detection module is used to reconstruct a three-dimensional model of the power cable sample based on the preprocessed power cable sample image, and perform three-dimensional defect detection to obtain a three-dimensional defect detection result of the power cable sample; A defect detection fusion module, which is used to fuse and analyze the two-dimensional defect detection results and the three-dimensional defect detection results of the power cable sample to generate a final defect detection result of the power cable sample; The step of fusing and analyzing the two-dimensional defect detection result and the three-dimensional defect detection result of the power cable sample to generate the final defect detection result of the power cable sample comprises: aligning the two-dimensional defect coordinate system in the two-dimensional defect detection result with the three-dimensional defect coordinate system in the three-dimensional defect detection result, generating a projection relationship expression, and performing fusing and analysis based on the projection relationship expression; After aligning the two-dimensional defect coordinate system in the two-dimensional defect detection result with the three-dimensional defect coordinate system in the three-dimensional defect detection result, generating a projection relationship expression, and performing fusion analysis based on the projection relationship expression, including: when a two-dimensional defect and a three-dimensional defect exist at the same coordinate position in the projection relationship expression after the two-dimensional defect coordinate system is aligned with the three-dimensional defect coordinate system, it is determined that there is a defect at this coordinate position.

7. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for detecting defects in a power cable according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for detecting defects in a power cable according to any one of claims 1 to 5 is implemented.

Images