ADSS optical cable electro-corrosion target detection method based on incidence matrix

Abstract

The invention provides an ADSS optical cable electro-corrosion target detection method based on an incidence matrix, and relates to the field of electric power detection combined with computer vision. The method comprises the following steps: acquiring current image data of an ADSS optical cable, analyzing the current image data, and extracting a plurality of electro-corrosion areas; extracting twisted wire-electrocorrosion key feature points in one of the electrocorrosion areas; constructing an incidence matrix based on the key feature points; and expanding the incidence matrix to a reference image, and positioning accurate positions of multiple electro-corrosion positions. According to the method, innovation and optimization are carried out aiming at the problems of varied electro-corrosion forms, high omission ratio and the like of the ADSS optical cable, and a detection scheme combining a target detection algorithm and an image processing technology is provided. According to the method, the incidence matrix technology is combined, the electro-corrosion feature points are accurately extracted and the matching matrix is generated through the feature point matching and amplification algorithm, and the accurate positioning of the electro-corrosion position of the optical cable is realized by utilizing the scene priori knowledge that electro-corrosion is mainly concentrated near the twisted wire.

Description

Technical Field

[0001] The present invention relates to the field of electric power detection combined with computer vision, and in particular to an ADSS optical cable electric corrosion target detection method based on a correlation matrix. Background Art

[0002] ADSS optical cable (All-Dielectric Self-Supporting Optical Fiber Cable) has become one of the important applications in the field of power communication due to its all-dielectric self-supporting characteristics. However, despite its superior weather resistance and corrosion resistance, which makes it more suitable for long-term field installation, ADSS optical cable is particularly susceptible to electrical corrosion in certain areas. Among them, the twisted wire area of ​​the optical cable has become the main location for high incidence of electrical corrosion due to its structural characteristics and the influence of the external environment. The twisted wire area not only often accumulates moisture and pollutants, but also its complex structure is prone to potential difference, which accelerates the corrosion process under certain conditions. As the electrical corrosion effect intensifies, it is easy to cause the surface sheath of the optical cable to break, causing cable breakage failure.

[0003] The detection methods for ADSS cable electrical corrosion problems cover a variety of technical means, including optical microscope detection, chemical reagent detection, potential difference measurement, infrared thermal imaging detection, sound wave detection, and automatic detection methods based on image processing, etc., but their accuracy is relatively limited. The current detection methods mainly rely on manual inspections, which have problems such as low efficiency and large subjective errors.

[0004] There is an urgent need to accurately detect the electrical corrosion that may occur on the surface of ADSS optical cables based on computer vision technology, especially to effectively monitor the corrosion in the area near the twisted wires, so as to achieve comprehensive monitoring and early warning of the optical cable channel. Summary of the invention

[0005] Purpose of the invention: To propose an ADSS optical cable electrical corrosion target detection method based on correlation matrix to solve the above-mentioned problems existing in the prior art.

[0006] The present invention proposes a method for detecting ADSS optical cable electrical corrosion targets based on a correlation matrix, and the steps are as follows:

[0007] Acquire current image data of the target ADSS optical cable;

[0008] Analyze the current image data to preliminarily identify and extract several electro-corrosion areas of the ADSS optical cable;

[0009] Extract the twisted wire-electrocorrosion key feature points in one of the electrocorrosion areas;

[0010] Constructing a correlation matrix based on the extracted twisted wire-electrocorrosion key feature points;

[0011] The correlation matrix is ​​extended to a reference image to locate the precise positions of multiple electro-corrosion areas; the reference image includes image data of other electro-corrosion areas or other ADSS optical cables in the current image data.

[0012] In a further embodiment, the electrocorrosion area of ​​the ADSS cable is initially identified and extracted to detect the frame express:

[0013]

[0014] The detection frame The coordinates of the four vertices , , , Expressed in homogeneous coordinate form:

[0015]

[0016]

[0017]

[0018]

[0019] Where x and y are the horizontal and vertical coordinates of the center point of the detection frame, respectively, and w and h are the width and height of the detection frame, respectively.

[0020] In a further embodiment, in the identified electro-corrosion area, the twisted wire-electro-corrosion key feature points are extracted to form a feature point set , It represents the key feature points of twisted wire-electrocorrosion extracted from the current image data, which are expressed in matrix form as follows:

[0021]

[0022] In the formula, , Respectively represent the horizontal and vertical coordinates of the twisted wire-electric corrosion key feature points extracted from the current image data;

[0023] The descriptor of the key feature points of twisted wire-electrocorrosion satisfies the difference Gaussian function :

[0024]

[0025] In the formula, represents Gaussian blur, is the input image, k is the scale factor, is the scale parameter.

[0026] In a further embodiment, using a feature point set , generate the correlation matrix H, which represents the spatial mapping relationship between the current image data and the reference image data. The correlation matrix H is a 3×3 matrix, expressed as:

[0027]

[0028] According to the feature point matching relationship, construct the twisted wire-electrocorrosion key feature points extracted from the current image data Key characteristic points of the extended wire-electrocorrosion The system of equations between:

[0029]

[0030] Expanding the system of equations yields:

[0031]

[0032] In the formula, , Represent the key characteristic points of the expanded twisted wire-electrocorrosion The horizontal and vertical coordinates.

[0033] In a further embodiment, the correlation matrix is ​​extended to the reference image to locate the precise locations of multiple electrical corrosions, and the expression is as follows:

[0034]

[0035] In the formula, Indicates the electrical corrosion area detection frame identified by the current image data. Represents the detection box expanded on the reference image.

[0036] In a further embodiment, the present application optimizes the parameters of the association matrix H by minimizing the error, and the optimization formula is as follows:

[0037]

[0038] In the formula, and are the horizontal coordinates of the corresponding feature points in the current image and the reference image respectively, and n is the number of iterations.

[0039] In a further embodiment, the correlation matrix H uses the following iterative formula to iteratively expand the feature point pairs multiple times:

[0040]

[0041] In the formula, is the incidence matrix at the kth iteration, is the incidence matrix at the k+1th iteration, is the iterative adjustment amount.

[0042] Iteration adjustment amount The calculation of is as follows:

[0043]

[0044] Where α is the learning rate; E(*) is the error function;

[0045]

[0046] In the formula, is the weight coefficient.

[0047] In addition, the present invention also proposes an electronic device, which includes a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the ADSS optical cable electrical corrosion target detection method based on the association matrix disclosed in the first aspect.

[0048] In addition, the present invention also proposes a computer-readable storage medium, which stores at least one executable instruction. When the executable instruction is run on an electronic device, the electronic device executes the ADSS optical cable electrical corrosion target detection method based on the association matrix disclosed in the first aspect.

[0049] Compared with the prior art, the present invention has at least the following beneficial effects:

[0050] (1) Improve the precision and accuracy of electrical corrosion detection: The present invention can automatically identify and accurately locate electrical corrosion on optical cables by combining target detection and correlation matrix technology. The correlation matrix is ​​used to match and expand feature points between multiple images, which makes the detection range wider, significantly improves the detection accuracy, and reduces the occurrence of missed detection and false detection.

[0051] (2) Reducing inspection costs and improving operational efficiency: Compared with traditional manual inspections, the present invention can significantly reduce inspection costs through intelligent monitoring methods, cover multiple key points and time periods of optical cable operation, and automatically generate monitoring reports and alarm information, greatly improving the efficiency of optical cable maintenance and detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 The present invention is a flow chart of the ADSS optical cable electrical corrosion target detection method based on the association matrix.

[0053] Figure 2 Schematic diagram of electrical corrosion in the stranded wire area of ​​ADSS optical cable.

[0054] Figure 3 Schematic diagram of SIFT feature point extraction for ADSS optical cable electrical corrosion. DETAILED DESCRIPTION

[0055] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features known in the art are not described.

[0056] The voltage-equalizing ring is an optical cable accessory used to equalize the voltage distribution of the optical cable in the electric field. Although the widespread use of insulating rings can reduce the probability of electrical corrosion, with the increase in the operating time of the ADSS optical cable, the accumulation of electrical marks, the effects of rain and dirt, and the influence of its local discharge effect, it is easy to cause the outer sheath of the optical cable to heat up, deform and be damaged, and the corrosion process to gradually accelerate.

[0057] Therefore, this embodiment aims to propose a method for detecting ADSS optical cable electrical corrosion targets based on a correlation matrix, and accurately detect the electrical corrosion that may occur on the surface of the ADSS optical cable based on computer vision technology, especially to effectively monitor the corrosion situation in the area near the twisted wire, so as to achieve comprehensive monitoring and early warning of the optical cable channel. The process is shown in Figure 1 As shown, the implementation steps are as follows:

[0058] Step 1: Image data acquisition

[0059] Step 1a: Image acquisition

[0060] The system regularly captures real-time images of ADSS optical cables through cameras installed in the optical cable channel. These cameras should have high resolution and the ability to adapt to all-weather environments to ensure that they can capture subtle electrical corrosion phenomena on the surface of the optical cable.

[0061] Step 2: Target detection and electrical corrosion identification

[0062] Step 2a: Model loading

[0063] The system analyzes the acquired images through the pre-trained YOLO model. The training data set of the model contains various forms of electrical corrosion phenomena (points, lines, cracks, etc.). The model can efficiently detect the electrical corrosion area and provide the location information and confidence of each detection frame.

[0064] Step 2b: Object Detection Output

[0065] After being processed by the model, the system outputs the location information (such as bounding box coordinates) and confidence score of each detected electrocorrosion phenomenon, which is used to indicate the probability of electrocorrosion phenomenon. These detection results provide preliminary electrocorrosion areas:

[0066] Detection box : The rectangular box output by the model represents the corrosion area. , where x and y are the coordinates of the center point of the detection box, and w and h are the width and height respectively.

[0067] Next, the coordinates of the four vertices of the detection box are expressed as homogeneous coordinates:

[0068]

[0069]

[0070]

[0071]

[0072] Step 3: Feature point extraction and correlation matrix construction

[0073] Step 3a: Feature point extraction

[0074] In the identified electro-corrosion area, the local feature extraction algorithm is further used to extract the twisted wire-electro-corrosion key feature points. The selection of feature points is not only based on the local features of the image, but also limited by the output corrosion area frame to reduce the interference of irrelevant information. In the area, the SIFT (Scale Invariant Feature Transform) algorithm is used to extract key feature points to form a feature point set .

[0075] Step 3b: Construct the incidence matrix H

[0076] Using the extracted twisted wire feature points, the correlation matrix H is constructed:

[0077] First, the feature points are represented by homogeneous coordinates:

[0078]

[0079] in, and are the corresponding feature points in the current image and the reference image respectively.

[0080] For perspective transformation on a two-dimensional plane, the correlation matrix H is a 3×3 matrix, expressed as:

[0081]

[0082] According to the matching relationship of feature points, a linear equation system is constructed:

[0083]

[0084] Expand to get:

[0085]

[0086] Since the coordinates after feature matching are still homogeneous coordinates, they need to be normalized:

[0087]

[0088] Step 4: Use the correlation matrix to distinguish the skeins from other parts

[0089] Step 4a: Through the characteristics of the correlation matrix H, the twisted wire area can be effectively distinguished from other areas:

[0090] Analyze the matrix eigenvalues ​​and calculate the eigenvalues ​​and eigenvectors of the correlation matrix. The characteristic value of , the corresponding eigenvector represents the symmetric transformation.

[0091] Define the discriminant function :

[0092]

[0093] like If it is close to 0, it is judged as a twisted wire area; If it is greater than a certain threshold, it is determined as a non-twisted wire area.

[0094] During the feature point matching process, the matching error is calculated:

[0095]

[0096] In the twisted wire area, the error is smaller; in other areas, the error is larger.

[0097] Step 5: Precise location of the electrocorrosion area

[0098] Using the characteristics of the correlation matrix, the precise location of the electrocorrosion area can be achieved:

[0099] Map the electrocorrosion areas in the current image to the reference image:

[0100]

[0101] in, is the bounding box of the electrocorrosion area, and H is the association matrix. The corresponding vertex coordinates are:

[0102]

[0103] When the correlation matrix changes abnormally in the twisted wire area (such as the matrix elements deviate from ), which may indicate electrical corrosion or other anomalies in the area and require further inspection.

[0104] Step 6: Iterative optimization of the incidence matrix

[0105] In order to improve the accuracy of the correlation matrix, an iterative optimization method is used:

[0106] The correlation matrix of the twisted wire area as initial value.

[0107] Iteration formula:

[0108]

[0109] in, is the adjustment amount for the kth iteration.

[0110] Adjustment calculation: , α is the learning rate; E(*) is the error function:

[0111]

[0112] in, is the weight coefficient, and the weight of the twisted wire area is higher.

[0113] When the error Converge to threshold or the maximum number of iterations is reached When , stop the iteration.

[0114] Below is an example of the correlation matrix comparing the twisted and non-twisted regions:

[0115] After calculation, the correlation matrix of the twisted wire area is:

[0116]

[0117] The diagonal elements of this matrix are close to , the off-diagonal elements are close to 0.

[0118] In the non-twisted wire region, the correlation matrix is:

[0119]

[0120] The elements of this matrix do not show obvious characteristics, and the diagonal elements are close to 1.

[0121] Use the characteristics of the correlation matrix to distinguish and calculate the discriminant function:

[0122] For the twisted wire area:

[0123] For non-twisted areas:

[0124] Setting Thresholds ,but:

[0125] like , determined as the twisted wire area;

[0126] like , determined as a non-twisted wire area.

[0127] result:

[0128] Twisted wire area , which is determined as the twisted wire area; , determined as a non-twisted wire area.

[0129] In the next step, the identified twisted wire area is used for target detection of electrocorrosion, which can effectively locate and identify the electrocorrosion area. The correlation matrix of the twisted wire area has stable characteristics and can be used as a benchmark; the correlation matrix of the electrocorrosion area is abnormal due to structural deformation. By using the numerical difference of matrix elements and the increase of matching error, combined with eigenvalue analysis, the electrocorrosion area can be accurately identified and its position can be accurately located through matrix operations.

[0130] The ADSS optical cable electrical corrosion target detection method based on the association matrix disclosed in the above embodiment can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiment can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs.

[0131] When the computer instruction or computer program is loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instruction may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instruction may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0132] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0133] Through the above specific implementation methods, the present invention can realize the automatic and accurate detection of optical cable electrical corrosion through the improved target detection algorithm and image feature point matching technology, solving the problems of low efficiency and high missed detection rate of traditional manual inspection. At the same time, the present invention improves the accuracy and coverage of optical cable electrical corrosion detection by introducing correlation matrix technology, ensuring the safety and stability of power communication.

[0134] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes may be made to it in form and detail without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A method for detecting ADSS optical cable electrical corrosion targets based on correlation matrix, characterized in that: The steps include: Acquire current image data of the target ADSS optical cable; Analyze the current image data to preliminarily identify and extract several electro-corrosion areas of the ADSS optical cable; Extract the twisted wire-electrocorrosion key feature points in one of the electrocorrosion areas; Constructing a correlation matrix based on the extracted twisted wire-electrocorrosion key feature points; The correlation matrix is ​​extended to a reference image to locate the precise positions of multiple electro-corrosion areas; the reference image includes image data of other electro-corrosion areas or other ADSS optical cables in the current image data.

2. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 1 is characterized in that: Preliminary identification and extraction of the electrical corrosion area of ​​the ADSS optical cable to detect the frame express: ； The detection frame The coordinates of the four vertices , , , Expressed in homogeneous coordinate form: ； ； ； ； Where x and y are the horizontal and vertical coordinates of the center point of the detection frame, respectively, and w and h are the width and height of the detection frame, respectively.

3. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 2 is characterized in that: In the identified electro-corrosion area, extract the twisted wire-electro-corrosion key feature points to form a feature point set , It represents the key feature points of twisted wire-electrocorrosion extracted from the current image data, which are expressed in matrix form as follows: ； In the formula, , Respectively represent the horizontal and vertical coordinates of the twisted wire-electric corrosion key feature points extracted from the current image data; The descriptor of the key feature points of twisted wire-electrocorrosion satisfies the difference Gaussian function : ； In the formula, represents Gaussian blur, is the input image, k is the scale factor, is the scale parameter.

4. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 3 is characterized in that: Using feature point sets , generate the correlation matrix H, which represents the spatial mapping relationship between the current image data and the reference image data. The correlation matrix H is a 3×3 matrix, expressed as: ； According to the feature point matching relationship, construct the twisted wire-electrocorrosion key feature points extracted from the current image data Key characteristic points of the extended wire-electrocorrosion The system of equations between: ； Expanding the system of equations yields: ； In the formula, , Represent the key characteristic points of the expanded twisted wire-electrocorrosion The horizontal and vertical coordinates.

5. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 4 is characterized in that: The correlation matrix is ​​extended to the reference image to locate the precise locations of multiple electrical corrosions. The expression is as follows: ； In the formula, Indicates the electrical corrosion area detection frame identified by the current image data. Represents the detection box expanded on the reference image.

6. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 4 is characterized in that: Also includes: The parameters of the correlation matrix H are optimized by minimizing the error. The optimization formula is as follows: ； In the formula, and are the horizontal coordinates of the corresponding feature points in the current image and the reference image respectively, and n is the number of iterations.

7. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 6 is characterized in that: The association matrix H uses the following iterative formula to iterate and expand the feature point pairs multiple times: ； In the formula, is the incidence matrix at the kth iteration, is the incidence matrix at the k+1th iteration, is the iterative adjustment amount.

8. The ADSS optical cable electrical corrosion target detection method based on correlation matrix according to claim 7 is characterized in that: Iteration adjustment amount The calculation of is as follows: ； Where α is the learning rate; E(*) is the error function; ； In the formula, is the weight coefficient.

9. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the ADSS optical cable electrical corrosion target detection method based on the association matrix as described in any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: The storage medium stores at least one executable instruction. When the executable instruction is executed on the electronic device, the electronic device executes the ADSS optical cable electrical corrosion target detection method based on the association matrix as described in any one of claims 1 to 8.

Images