Effective scheme of measuring the displacement between wheel and rail of active steering bogie using computer vision system

Abstract

This paper proposes an effective scheme of measuring the displacement between wheel and rail of active steering bogie using computer vision system for overcoming problems of noise and safety. The paper presents a monitoring system, which can supervise the condition of rail and active steering bogie's running condition. Nowadays, the rail transportation is an important for city conurbation. But steering for railway vehicles has problems with noise and safety. Measuring the displacement between wheel and rail of active steering bogie prevents accident in the curve section and intersection section. This paper employs template matching algorithm and projection for detection of wheel. In the case of rail, the labeling algorithm is used to get rail candidates from a lot of edge components in the rail region. The edge components are consisting of important ingredients and noise. It abstracts important edge components from grouped edge components. Finally, detection of rail is applied to adaptive Hough transform algorithm. The displacements are calculated in world coordinate. So displacements of Image coordinate is transformed to world coordinate. The proposed algorithm is performed with various wheel-rail images on outdoor, intersection, curve and tunnel railroads. This helps to describe and compare characters of measurement in the various experiment images. © 2009 SICE.