Research On Image Recognition And Detection Of Overhead Transmission Line Breaking

Transmission line plays an important role in the transmission of electrical energy to all parts of the country in the power system,and are also the basis for the safe and stable operation of the power system.Transmission line are vulnerable to external influences in the natural environment for a long time.The broken strands will have great harm to the safe operation of transmission line.In order to ensure the safe operation of the line,it is necessary to strengthen the inspection and maintenance of the line so as to eliminate hidden dangers in time.At present,the inspection method has gradually changed from manual inspection to drone inspection.In order to improve the efficiency of transmission line inspection,image processing technology is applied to the UAV line.This article has mainly completed the following work:1.Aiming at the problem that the transmission line image is susceptible to illumination,the histogram equalization,gamma transformation and piecewise linear image enhancement methods are studied.Finally,the histogram equalization combined with gamma transformation is used to enhance the image.This method retains the image.The details of the power line reduce the appearance of false contours and improve the contrast of the power line image.For the noise problem generated when the UAV collects images,the image is denoised by mean filtering,median filtering and Gaussian filtering respectively.Finally,median filtering is used to denoise,which improves the image quality.2.Aiming at the threshold selection problem of binary image of power transmission line,by comparing various threshold segmentation algorithms,it is concluded that the maximum inter-class variance algorithm can better distinguish the transmission line from the background.Then the Canny edge detection is applied to the binarized image of the power line to extract a smooth,continuous line edge.For the edge of Canny’s extraction,there is a breakpoint problem,and then the Hough transform combined with the contour extractionmethod is used to improve the power line image,and the complete power line contour can be accurately extracted.Since the contour of the broken power line is larger than the contour of the normal power line,the contour of the power line with the fork is extracted by finding the maximum contour.Finally,using the closed operation in morphology,the broken power line is extracted.3.Aiming at the characteristics of the broken wire of the transmission line,a method for detecting the broken wire of transmission line according to the end point of transmission line is proposed.The method calculates a distance between the bifurcation end point and the normal power transmission line,sets a threshold value,and judges whether the break point is a fault point according to the distance and the threshold value.This method can quickly realize the fault identification of the power line breakage.4.An automatic detection system for the breakage of transmission lines is designed.In this system,the image of the transmission line collected by the UAV is tested and analyzed according to the test results.The test results show that the system can accurately detect the position of the broken wire of the power line.