Research On Track Crack Detection Technology Based On Image Processing

With the rapid development of high-speed railways today,CRTS? type slab track is the longest kind of ballastless track structure in China's current 350km/h high-speed railway.Rail safety involves the lives and safety of the people.The track safety issue of CRTS? type slab track has become a hot topic.Track plate cracks are one of the key hidden dangers in track safety.Because the environment is complicated and the image acquisition is difficult,the crack detection of the track plate is a difficult point.This paper is based on a comprehensive analysis of research status at home and abroad.For the CRTS? type slab track,the track crack detection technology based on image processing has been fully studied.Main tasks as follows: A method for extracting significant areas of cracks in track plates based on image processing in a complex environment is proposed.Track crack images collected in a complex environment have problems such as uneven illumination and high noise.At first,image preprocessing is performed including scale transformation,Gaussian filtering,and histogram equalization.Then,the K-means algorithm and the open operation are used to realize the rough classification of the track crack image.Finally,the Canny edge detection algorithm is improved.The algorithm extracts a clear area of sharp cracks.A method of track crack detection and classification based on image processing is proposed.Firstly,the crack feature is extracted using an algorithm based on the shape of the crack.Then,the exact location of the crack is made and the crack in the image is marked with a minimum circumscribed rectangle.Finally,a standard for classifying crack disease grades is proposed.The algorithm effectively improves the detection efficiency of crack detection on track plates in complex environments.The algorithm accuracy rate is 87.73%.The grade of the crack disease was quantified.