The Design Of Rail Light Image Detection System

High-speed railway security detection has become one of the important tasks of our country the development of high-speed railway,guarantee the safe and efficient operation of high-speed train is a major challenge for the development of our countryhigh-speed railway.Light rail is one of the important indicators of the research of track irregularity,according to the distribution of light rail can detect and judge smooth surface of the rail.When steel band exception occurs,the direct reason is that the wheel/rail interactions and point has changed.Detection of track irregularitiescan starts from the first light.In order to find an efficient,automatic and intelligent rail stationarity test,the method of artificial detection of manpower,low detection efficiency and accuracy is not sufficiently protect the shortcomings,this paper builds the rail light detection system,it uses the machine vision technology and computer technology,it can effectively reduce the manpower,improve the efficiency of detection,and to ensure theaccuracy of detection.This paper mainly researches content of the steel band test system of band structures,rail test system of image analysis,light rail defects analysis and the analysis of the code migration,etc.For the construction of the rail light detection system,we start from hardware and software,including the selection and installation of the CCD camera,image acquisition software and calibration blocks,etc.Steel band image analysis aspects of detection system,this paper has carried on the calibration,image preprocessing,image processing technology such as morphology,recovery.For the light rail defect analysis,mainly in image restoration,image edge detection and image histogram is analyzed.The code of system analysis,two alternatives are presented in this paper,the MATLAB Coder and the MATLAB Compiler.Based on the characteristics of light rail we choose the CCD camera and verified its feasibility from the angle of error analysis.For the detection of rail light can use first image calibration,average filtering and histogram equalization enhancement preprocessing,and then morphological image processing,the rail light detection error analysis,finally combining blind deconvolution image restoration are analyzed.In terms of light rail defect analysis,using blind deconvolution image restoration algorithm,canny algorithms of image edge detection and image histogram.Transplantation for system code,can be selected according to actual demand different solutions.By analyzing,the surface of CCD camera satisfy the system requirements,rail light detection effect is more ideal,blind deconvolution image restoration algorithm rail band of the motion blur effect is good.In terms of light rail defect analysis,can be combined with image edge detection and image histogram infer that rail light if there is a defect.Transplantation for system code,can according to actual needs and combined with MATLAB Coder and MATLAB Compiler features to choose C/C + + code generation system.