Underbody component monitoring system of railway vehicles using the infra-red thermal images

Abstract

This paper is on the study of the status monitoring system using infra-red camera for the key underbody components of railway vehicles (traction motors and reducers). Image processing and fault diagnosis algorithms that enable the efficient detection of abnormality of the key underbody components of railway vehicles are important factors for safe railway operations. For the field tests, underbody images of the KTX trains have been collected as they enter the depot. By calculating the position of the components with the reference to the maximum temperature of the targets, the abnormal overheat is detected using template matching technique. Moreover, advanced methods have been added to the conventional template matching process for the fast image processing using high speed camera such as minimum temperature setting for the region, pattern size designation, and discarding repeated images. And in order to improve the detection rate, the templates with different suitability in the same frame are compared using AND calculation.