Research On Identification Method Of Status High Voltage Transmission Lines

High-voltage transmission line is the key equipment of the power system.Maintaining its normal operation is an important prerequisite for the stable operation of the power grid.The image processing technology is used to realize the identification of the dangerous state of the high-voltage transmission line.So as to timely detect the potential safety hazards of the high-voltage transmission line,which is of great significance to the normal operation of the power system.Aiming at the defect that the traditional Canny edge detection algorithm misses the weak edge and relies on the experience to set the double threshold,this paper proposes an adaptive Canny edge detection algorithm,which is used to complete the edge detection of high-voltage line image.In order to retain more image details,the algorithm firstly smooth the image through the hybrid filter which is composed of the geometric mean filter and the median filter.Then,by increasing the edge gray value by linear edge enhancement,a set of candidate foreground pixels and candidate background pixels are obtained.Finally,the optimal threshold calculated by the improved Otsu algorithm replaces the high threshold of the double threshold algorithm for edge detection and connection.Thereby eliminating the limitation that the Canny algorithm must manually set the double threshold and extracting the edge details of the image well.Aiming at the problem that the traditional Freeman chain code algorithm has a broken line and a meaningless line on the wire extraction,this paper proposes a high-voltage transmission line extraction algorithm based on Freeman chain code.The algorithm is based on the problem that the traditional algorithm is limited by the initial search direction and the neighborhood search range,which leads to the occurrence of straight line breaks and meaningless lines.Firstly the Hessian matrix is used to calculate the feature vector of each pixel of the original image,and then the statistics of the high-voltage line in the image are used to obtain the main direction of the high-voltage line.Then,according to the obtained main direction,the initial search direction and search mode of the Freeman chain code are determined,and the calculation amount and the meaningless line appear to reduce the line segment corresponding to the main direction and fit.Finally extracting the high-voltage transmission wire by straight-line combination.The dangerous state of high-voltage transmission lines mainly includes foreign body,ice coating and broken strands.In this paper,a dangerous state identification algorithm for high-voltage transmission lines based on Harris corner is proposed.Firstly,the initial corner set is obtained by Harris corner detection preprocessing,then the pseudo corner points are filtered out by USAN and the corner points not on the high-voltage line are removed according to the direction of the high voltage line,thereby obtaining the abnormal corner point set.Secondly,the analysis is based on the variation of the width of the three dangerous states of the wire.For the high-pressure line foreign body state,the high-voltage line will exhibit a width change in the foreign body range,and the width change rate and the change area are large.For the high-voltage line ice coating state,there are mainly two types.One type of high-voltage line is covered with ice,and the overall width of the transmission line is larger than that of the normal line,but the width change rate remains basically unchanged.The other is that the high-voltage line is partially covered with ice.In this case,the width of the high-voltage line will locally become larger,and the changed area is larger.The width of the wire in the ice-covered part is significantly different from that of the uncovered wire.For the high-voltage line breaking state,the width of the high-voltage line will appear smaller than the normal wire,the width of the broken strand will vary greatly,and the rate of change of the rest will be stable.Therefore,using the variation of the width of different dangerous states,the abnormal corner points are used as the starting point for identification,and finally the dangerous state areas are marked.