Railway Foreign Object Detection Based On Digital Image Processing

Railway career in China has developed rapidly in recent years,a large number of railway lines are complete,have also made great progress and high-speed breakthroughs,it is convenient for people to travel,also put forward higher requirements on the safety of railway operation.For a long time,railway has been threatening the safety operation of railway foreign body problem.Rely on the traditional detection method and the driver visual for railway foreign body recognition difficult to thoroughly solve the problem.So,the use of machine vision,design a kind of effective railway foreign body detection system,can accurately detect the foreign body,real-time and sound an alarm in time,this will help ensure railway operation safety.In this paper,after studying the existing railway foreign body detection technology at home and abroad,designed a railway foreign body detection system based on digital image processing,mainly including railway foreign body detection system overall structure design and algorithm design,and through the simulation experiment proves that the feasibility and practicability of the design system in this paper.Railway foreign object detection based on digital image processing methods,design the detection system based on background subtraction algorithm,the basis of foreign matter detecting function is realized.On this basis,this paper analyzes the railway monitoring video image due to the change of acquisition environment led to the decrease of the quality of the image,such as light intensity change,wind,train passes the vibrations will cause the distortion camera,video camera dithering,make the video image is often contains less noise,contrast,image distortion is insufficient,affect after foreign body detection.In order to solve these problems,in this paper,the video image preprocessing:the use of image filtering,gray image contrast enhancement of image is optimized,respectively choose to apply a time domain recursive low pass filtering method,histogram equalization method.Then jitter analysis was carried out on the video image,such as a frame exists dithering,use the image registration algorithm based on corner feature to its registration.Given in this paper,image preprocessing method can effectively improve the quality of railway image,reduce the detection error rate,effectively improve the accuracy and reliability.In order to verify the practicability of this design system,the use of some relevant railway video moving object images and video,respectively in the MATLAB simulation platform for time domain recursive low pass filtering method,histogram equalization,image registration algorithm based on corner features and background difference method,the experimental results show that the algorithm used in this paper can effectively achieve the design purpose,quickly and accurately detect the foreign body,and can effectively distinguish foreign material properties,high detection accuracy,good real-time performance,suitable for railway foreign body detection.