Identification And Location Of Catenary Insulator Based On Binocular Vision

With the rapid building of modern railway network in our country,the safe operation of the rail system is required more strictly.As an important component of electrified railway,ensuring the safe operation of power supply system is more and more important.The large area blackout caused by the flashover not only affects the normal railway operation,but also brings great loss to the economy.Maintain the cleanness of catenary insulator is an effective method to prevent flashover.At present,water washing cars flush the catenary insulators by manual operation.With the high intensity cleaning labor and low flushing efficiency,the catenaiy insulator is not clean if the workers do not operate properly.Automatic flushing for water washing car is an effective way to solve the problem of manual washing.In order to solve the problem of heavy insulator cleaning along the railway line,the use of automation equipment to replace the manual washing has become the focus of research.Catenary insulator's 3D reconstruction is the key of flushing automatically.In this paper,catenary insulator is the research object.our purpose is to extraction and location the catenary insulator so that restore the three-dimensional space position by using binocular vision technology.The feasibility of the method is verified by experiments,which provides the basis for the automatic flushing of the contact catenary insulator.The main contents of this paper are as follows:1.This paper presents a new method for the identification of catenary insulators based on color and correlation.According to the color characteristics of insulator,the approximate insulator region is extracted,then the insulator is identified according to the relative position between the contact wire and the insulator.2.Using the method of plane calibration template and square chessboard as the 2D plane target to calibrate the binocular vision system in MATLAB,so that internal and external parameters of the camera are obtained,and it establish a connection between the image coordinates to world coordinates for three-dimensional reconstruction of catenary insulator.3.Studying the corresponding point matching technique in stereo matching.According to the matching accuracy and real-time requirements,analysis the characteristics of catenary insulator,Choose the combining method by centroid matching and gray information matching which is applicable for insulator images.4.Analyze the principle of 3D vision reconstruction of binocular vision system with parallel optical axis and camera position variable.The experimental platform of binocular vision system is built,and the insulator image is captured in the 1200-3200mm range to restore 3D position information of insulator.The reconstruction error and the real-time analysis are analyzed to verify the feasibility of 3D reconstruction method.