The Research And Realization Of Foreign Matter Detection For High-speed Railway Catenary

China’s high-speed railway has been developing rapidly in recent years,its influence on public life is getting more and more widely.The safe operation of high-speed railway is related to the health and life of passengers.So it is particularly important.The high-speed railway catenary is the device which supplies power to electric locomotives.Because foreign matter may be attached to this device,the safety of high-speed railway is threatened.At present,catenary security inspection device takes pictures of high-speed catenary firstly.Then inspectors observe these images to judge whether foreign matter is attached to the catenary or not.This kind of detection spends much time and consumes a lot of energy.To improve the speed of detection,this paper studies the high-speed railway catenary foreign matter detection technology,and realizes the high-speed railway catenary foreign matter detection system.The foreign matter which is attached to the high-speed railway pillar is the focus of this paper.On the basis of the existing catenary security inspection device,the paper realized the automatic detection technology of high-speed rail catenary image.This paper uses machine learning and image processing technology to achieve detection algorithm,and the paper builds the system with MFC framework.The system completed some function,including intelligent preliminary detection,artificial fine detection,log information recording and querying,user information managing.Intelligent preliminary detection is core content of this paper.First step is extracting features of positive and negative sample images.These features are input data of SVM.Through training SVM to generate a high-iron catenary pillar classifier.Second step is using this classifier to identify the pillar area in images which to be detected.Third step is comparing similarity between pillar images and standard templates.By this way we can judge whether foreign matter is attached to the catenary or not.Combining the application environment,this paper proposed a new method in similarity comparison process.The paper detects edges of images firstly to preserve the structure information.After this process,the paper can get more accurate similarity comparison results.Artificial fine detection is the platform for inspectors to recheck images.The paper establishes this platform to pinpoint the catenary foreign matter precisely.The system also has log information recording and querying function and user information managing function.These two functions provide a convenient way to manage information.By this way, the system can improve the information level of detecting work.This paper mainly introduces the design and implementation of high-speed railway catenary foreign matter detection system.Firstly,this paper describes the background and purpose of the project.This paper also describes the research status of catenary detecting technology both at home and abroad.Then the paper introduces some relevant technology.After this,considering the specific circumstances of the project,the paper describes the system’s requirement analysis,system design and some realization of the function with.Finally,the work of this paper is summarized and the future work direction is put forward.