Detection Of Pantograph And Catenary Pole Based On Monitoring Video

In recent years,the high-speed railway in China is in a period of rapid development.The basic frame of "4 length lines and 4 breadth lines" high-speed railway network is formed and train speed is also increasing.Therefore,the safety and reliability of railway has been increasingly emphasized and the safe operation of power supply system plays a key role.In order to meet the increasing demand of railway power supply system security detection and monitoring,relieve the stress of manual testing detection and achieve the automatic and intelligent safety inspection of pantograph-catenary system,more and more attention has been paid to the detection and monitoring methods based on image processing technology.The research work of this paper is based on the Catenary-Checking Video Monitor Device and Pantograph Video Checking Device in 6C system.This paper takes the images of pantograph and catenary inspection as experimental data,realizes the intelligent detection and extraction of the target device in the image by means of image processing and machine learning.The effectiveness of the proposed algorithm in this paper is also demonstrated by experiments.The main work and innovation of this paper include the following contents:In the process of image preprocessing,the CLAHE algorithm is studied firstly,and is used to demist the images whose contrast is not obvious and which is influenced by fog.Then the method which combines Hough transform with Canny Algorithm is proposed to detect the angle of EMU roof image.The Rotating projection method is used to detect the angle of the catenary image.Finally,perspective transformation is applied to horizontal correction of two types of images.In the detection of pantograph,the combination of Sobel operator and morphological operation is used to extract the pantograph region roughly.Then,the Paralleled-Gabor transform is studied to extract the directional characteristics of the pantograph region.Finally,the decision fusion method based on multiple SVM classifiers is used to realize the detection of pantograph region accurately.In the detection of catenary pole,the acquisition of perspective information in catenary image and rail position adopted the vanishing point detection method.Afterward based on standards of high-speed railway design and construction,by calculating the catenary post and Sampling,the suspected post area can be obtained.And the convolution neural network is studied to be used for the detection and extraction of catenary post.Finally,the EMU roof image and the catenary inspection image are tested.The algorithm has good applicability and ideal recognition rate.It is proved that the algorithm has certain engineering application value.