Image analysis for detecting the transverse profile of worn-out rails

Abstract

Over its useful life a railway track is subject to
many mechanical and environmental stresses which gradually
lead to its deterioration. Monitoring the wear condition of the
railway superstructure is one of the key points to guarantee an
adequate safety level of the railway transport system; in this
field, the use of high-efficiency laser techniques has become
consolidated and implemented in diagnostic trains (e.g. the
“Archimede train” and the “Talete train” ) which allow to detect
the track geometric parameters (gauge, alignment, longitudinal
level, cross level, superelevation defect, etc.) and the state of rail
wear (vertical, horizontal, 45-degree etc.) with very high
accuracy. The objective of this paper is to describe a new nonconventional
procedure for detecting the transverse profile of
worn-out rails by means of image-processing technique. This
methodological approach is based on the analysis of the
information contained in high-resolution photographic images of
rails and on specific algorithms which allow to obtain the exact
geometric profile and the measurement of the relevant deviations
compared to new rails of the same typology.
The analyses and the first results, obtained from laboratory
researches, concern rails cross sections taken from railway lines
under upgrading. The procedure has shown high precision in the
wear evaluation as well as great rapidity in being performed.