Method of Real-time Recognizing Effective Rail Profiles from Complex Track Structures

Real-time and accurate recognition of the effective rail profiles from the original ones acquired by the onboard rail profile detection system to assess the quality of track is a key challenge to rail inspection. Firstly, on the basis of the comparative analysis of the rail profiles located in different areas of the railway, the discontinuity between rail head and rail waist and the match between measured rail waist and standard one were used as the important basis to recognize the effective profiles. Then, in view of the defects of the unknown status of rail waist of measured profile and insufficient feature points in other areas, a novel method using the rail jaw point and railhead inner line to construct the curve registration matrix was proposed. Last, the measured profile was classified precisely by computing the directed Hausdorff distance of rail waist overlap areas and comparing it with statistical threshold. The experiment results on actual railway show that the distributions of effective profiles based on the proposed method are accurately coincided with the real conditions, with the recognition accuracy rate reaching 93.3%, and the average recognition speed reaching 10.8ms per frame. The measured system can run at a speed of 83.3 km/h under the experimental setup, which has a definite theoretical and engineering application value. © 2020, Department of Journal of the China Railway Society. All right reserved.