CN-119206240-B - Power line identification method, system, equipment and storage medium

CN119206240BCN 119206240 BCN119206240 BCN 119206240BCN-119206240-B

Abstract

The invention provides a power line identification method, a system, equipment and a storage medium, which belong to the technical field of power system detection, wherein the method comprises the following steps of obtaining an image to be detected containing a power line; the method comprises the steps of constructing a phase filter stretching kernel function based on a phase translation kernel function and a convolution theorem, inputting an image to be detected into the phase filter stretching kernel function, stretching and transforming the image to be detected by using the phase filter stretching kernel function, converting the image to be detected in a frequency domain into an angle image in a spatial domain, extracting contour edges in the angle image, superposing the image to be detected and the angle image containing the contour edges to obtain a superposed image, sharpening and enhancing the superposed image by using relative total variation RTV, and identifying a power line. The invention can improve the accuracy of identifying the power line.

Inventors

XU PENGFEI
DING YAFEI
ZHAO JIANGWEI
CUI BING
WANG YUEYAN
WANG CHENYANG
Yan zijing
YUE XUETING

Assignees

平顶山学院

Dates

Publication Date: 20260508
Application Date: 20240904

Claims (6)

1. A power line identification method, characterized by comprising the steps of: acquiring an image to be detected containing a power line; Based on a phase translation kernel function and a convolution theorem, carrying out angle transformation on a space domain complex matrix corresponding to a complex stretching kernel of a frequency domain of the phase translation kernel function and an image after convolution to obtain an angle image after phase stretching, simplifying the angle image after phase stretching, and carrying out polar coordinate transformation and normalization processing on a nonlinear phase distortion kernel function to obtain a phase filter stretching kernel function with a phase stretching intensity parameter S and a distortion parameter W, wherein the phase filter stretching kernel function is Wherein S is a phase tensile strength parameter, W is a distortion parameter, and r is a polar diameter when converting polar coordinates; Inputting an image to be detected into a phase filter stretching kernel function, stretching and transforming the image to be detected by using the phase filter stretching kernel function, converting the image to be detected in a frequency domain into an angle image in a space domain, and extracting contour edges in the angle image; and superposing the image to be detected and the angle image containing the contour edge to obtain a superposed image, carrying out sharpening enhancement processing on the superposed image by using relative total variation RTV, and identifying the power line from the image subjected to the sharpening enhancement processing.
2. The method for identifying the power line according to claim 1, wherein before the image to be detected is input into the phase filter stretching kernel function, the method further comprises the steps of carrying out graying treatment and filtering treatment on the image to be detected, wherein the graying treatment is specifically carried out by taking a weighted average value of 3 values of an R component, a G component and a B component of a pixel in the image to be detected as a gray value of a gray map, and the filtering treatment is carried out by carrying out weighted average on all pixel values of the image to be detected, wherein the pixel value of each pixel point is obtained by carrying out weighted average on the pixel value of each pixel point and other pixel values in the neighborhood.
3. The power line identification method according to claim 1, wherein the angle image is: Wherein, the For the phase filter stretching kernel function, IFFT2 is the inverse of the two-dimensional fourier transform, The result is obtained by performing inverse Fourier transform processing on the input image B (x, y).
4. A power line identification system, comprising: The image acquisition module is used for acquiring an image to be detected containing the power line; The phase filter stretching kernel function is used for carrying out angle transformation after convolving a space domain complex matrix corresponding to a complex stretching kernel of a frequency domain of the phase translation kernel function with an image based on the phase translation kernel function and a convolution theorem to obtain an angle image after phase stretching, simplifying the angle image after phase stretching, carrying out polar coordinate transformation and normalization processing on a nonlinear phase distortion kernel function to obtain a phase filter stretching kernel function with a phase stretching intensity parameter S and a distortion parameter W, wherein the phase filter stretching kernel function is that Wherein S is a phase tensile strength parameter, W is a distortion parameter, and r is a polar diameter when converting polar coordinates; the contour extraction module is used for inputting the image to be detected into a phase filter stretching kernel function, stretching and transforming the image to be detected by using the phase filter stretching kernel function, converting the image to be detected in a frequency domain into an angle image in a spatial domain, and extracting contour edges in the angle image; and the power line identification module is used for superposing the image to be detected and the angle image containing the contour edge to obtain a superposed image, carrying out sharpening enhancement processing on the superposed image by using relative total variation RTV, and identifying the power line from the image subjected to the sharpening enhancement processing.
5. A computer device comprising a memory storing a computer program and a processor for running the computer program in the memory to perform the power line identification method of any one of claims 1-3.
6. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program adapted to be loaded by a processor for performing the power line identification method of any of claims 1-3.

Description

Power line identification method, system, equipment and storage medium Technical Field The invention belongs to the technical field of power system detection, and particularly relates to a power line identification method, a system, equipment and a storage medium. Background The power line inspection is an important component part of daily maintenance of the power grid, and plays a vital role in guaranteeing healthy operation of the power grid. However, a large amount of transmission lines are erected in mountain areas with inconvenient traffic, the manual line inspection efficiency is low, the defects can be overcome by using unmanned aerial vehicles to inspect the lines, but the definition of the real-time returned images is insufficient, and the missed inspection is easy to cause, so that the fault identification is deployed on the unmanned aerial vehicles to assist manual judgment, and the method is a good scheme. The segmentation of the power line from the image is the basis for performing the power line defect detection, and currently, the power line extraction can be classified into a conventional method and a data-based deep learning method. The method comprises the steps of obtaining an edge graph by a Canny operator and extracting a power line in the edge graph by Hough transformation, wherein the Canny algorithm has inaccurate excessively-effective conditions under the influence of noise when an image is segmented, and an improved anchor frame generation strategy is provided when the R-CNN (Region-Convolutional Neural Network) algorithm is utilized for detecting lines, but the R-CNN image is insensitive to the occurrence of image edges and has excessive errors such as undersegmentation when the R-CNN image is segmented, the phase stretching transformation (PHASE STRETCH transformation, PST) is developed on the basis of the time stretching transformation (TIME STRETCH transformation, TST) of an analog signal, the PST is used for occasions such as digital image edge feature detection, feature enhancement of a visual damage image and digital image compression, but the detected edges contain a large amount of noise, particularly fine-broken high-frequency noise points are mistakenly considered as high-frequency components in the image and are reserved, and the subsequent thresholding processing is difficult. In summary, since the power line target is tiny, the reflected energy is weak during imaging, so that the unmanned aerial vehicle aerial image is easily interfered by the environmental linear elements and noise, the power line imaging noise is serious, the imaging is weak and fuzzy, and the power line is not easy to distinguish, and therefore the effect of extracting the power line by the detection algorithms is not ideal. Disclosure of Invention In order to overcome the defect of non-ideal power line extraction effect, the invention provides a power line identification method, which comprises the following steps: acquiring an image to be detected containing a power line; Based on a phase translation kernel function and a convolution theorem, performing phase stretching transformation and convolution operation on the phase translation kernel function to construct a phase filter stretching kernel function; Inputting an image to be detected into a phase filter stretching kernel function, stretching and transforming the image to be detected by using the phase filter stretching kernel function, converting the image to be detected in a frequency domain into an angle image in a space domain, and extracting contour edges in the angle image; and superposing the image to be detected and the angle image containing the contour edge to obtain a superposed image, carrying out sharpening enhancement processing on the superposed image by using relative total variation RTV, and identifying the power line from the image subjected to the sharpening enhancement processing. Preferably, the construction of the phase filter stretching kernel function includes the following steps: carrying out angle transformation after convolving a space domain complex matrix corresponding to a complex stretching kernel of a frequency domain of the phase translation kernel function with the image to obtain an angle image after phase stretching; Simplifying the angle image after phase stretching, and carrying out polar coordinate conversion and normalization treatment on the nonlinear phase distortion kernel function to obtain a phase filter stretching kernel function with a phase stretching intensity parameter S and a distortion parameter W. Preferably, before the image to be detected is input into the phase filter stretching kernel function, the method further comprises the steps of carrying out graying treatment and filtering treatment on the image to be detected, wherein the graying treatment is specifically carried out by taking a weighted average value of 3 numerical values of an R component, a G component and a B component of a