Development of new methods for assessing the technical condition of power lines using digital image processing

Abstract

An important role in satisfying the high quality of life of the population and modern industry is played by the reliability of power supply and ensuring uninterrupted delivery of electricity. As a result of a large number of emergency failures of elements caused by external factors, information about the technical state of elements in real time is not used by dispatching centers. These conditions cause difficulties in assessing the reliability of power lines and timely elimination of developing defects, which cause high transmission losses and lead to wire breaks and failures.Considering a digital photo or other raster image to assess the surface of power lines wires, we know that it is an array of numbers recorded by sensors of brightness levels, in a two-dimensional plane. Knowing that in mathematical terms a thin lens performs the Fourier transform of images placed in the focal planes, it is possible to create image processing algorithms, which are analogues of image processing by a classical optical system. An algorithm for automatic processing of scanning electron microscopy images aimed at detecting and describing surface defects in power line wires is proposed, which is implemented in the Python programming language. The algorithm proposes to use a Fourier transform procedure to eliminate the image brightness gradient and suppress high-frequency noise in the image by applying a two-dimensional bandpass filter. The necessity and parameters of the dynamic range normalization of the filtered image are determined. The binarization parameters of the normalized image are determined. A quantitative assessment of the degree of defectiveness of the investigated wire surface has been proposed. A quantitative assessment of the degree of elongation of defects on the surface of the investigated wire has been proposed.