Defects in railroad tracks are responsible for several incidents every year. Rail gauge is one of the most important measurements for track maintenance, because deviations in gauge indicate where potential defects may exist. In addition, missing rail fasteners can be considered as a critical defect that should be detected and repaired as they are common cause of gauge misalignment issues. In this paper, an improvement or enhancement to currently available automatic inspection system specifically devised to estimate the rail gauge and detect missing rail fasteners is presented. A 3D imaging sensor, which produces high resolution 2D images and 3D profiles, is used to capture the data. Then a range-based approach is used to inspect the railroad track. We rely on the 3D structure of the rail components (rail heads and rail fasteners) instead of using a vision-based approach which suffers from illumination changes. The system is evaluated using data recorded from real scenarios in two different cities (Metro Madrid and London Underground), with different nominal gauge values and fastening elements. The system is described and results are presented, evaluated and discussed.
展开▼
