Research On The Technology Of UAV Power Line Inspection Based On Machine Vision

The power transmission lines have shouldered the important task in the process of power transmission, and Its operating states have closely bound up with people’s lives. Its regular inspection is an important measure to ensure the safe and stable operation of the power grid. But the traditional way of inspection inefficient, with the development of machine vision technology, it offers the possibility of automatic inspection, based on Machine Vision of UAV power line inspection techniques provide a feasible way, for the automatic inspection of power lines were also discussed aspects of the algorithm.This paper proposes a transmission line inspection used by UAV based on machine vision. According to the characteristics of a variety of common drones patrol the line, and analyze the strengths and weaknesses of various UAV patrol line. Choose fixed-wing UAV and quad-rotor UAV, or an unmanned helicopter, both economical and efficient way transmission line, then briefly introduces the structure of UAV drones patrol line system and the key issues. Aiming at the complex backgrounds of video images captured by UAV for inspection of transmission lines, and the question that image have noise and blur caused by mechanical shocks shooting, shot swing factors due to the high altitude, to obtain the maximum entropy by Wiener filter denoise and strike with the probability density function to blur the image preprocessing recovery mode, and improved clarity of the image. Finally, according to the power complex background, and used the improved Canny edge detection operator to achieve the initial image segmentation, so that the strength of the edge of the image can be a good testing, and suppress noise. For the problems of simultaneous appearing of transmission line transmission line contour extraction and background objects in the pre-image segmentation, and according to the characteristics of the transmission line, this paper proposed the transmission line extraction algorithm based on improved phase-grouping method and Hough transform, so that the transmission line is extracted from complex power background. This method also has the advantage of Hough Transform and phase-grouping method, and enhances the operating efficiency and real-time. Aiming at transmission Line for automatic tracking, and according to the existing tracking techniques and algorithms, this paper Proposed tracking algorithm based on the Kalman filter lossless (UKF) for Camshaft target tracking, and using the lossless Kalman filter (UKF) filtering method for predicting the transmission line identification of the region of interest, and adjust the forecast and the size of the search window by Camshaft algorithm to track real-time transmission lines. Finally, we realize the tracking algorithm MATLAB simulation, and verify that the tracking performance of the algorithm..