Detection of the Rail Profile Wear Based on Image Processing

For the rail head wear detection, the feasibility of collecting rail images by CCD camera is analyzed. MATLAB image processing technology is used to establish the calculation model of rail wear measurement. By using the methods of edge detection and image filling, the key problems of rail image positioning are studied. The results show that the rail abrasion image detection system can get the rail abrasion accurately in real time under the condition of non-contact. The method can calculate the rail wear more accurately and provide a reference for the loss prevention of the rail and the development of new detection technology.