Inspection Method for Track Components Defects based on Image Recognition in Subway System

Track components —such as rail, fastening, and ballastless slab — in urban rail transit are damaged frequently by train dynamic load or temperature and humidity of the environment around the track. Traditional inspection methods, like human inspection and identification, were low in efficiency and high in missing rate. In this paper, based on the theory of Image Recognition (IR), the identification algorithm for track components' condition and damages was researched. The image processing method, the identification operators and data transform method were analyzed. The image recognition method for track component damages was presented. The broken rail and crack in rail food from the photo graphics by inspection were recognized and identified. It shows this method has high accuracy and identification capability. It will be the key technology for the software of high-speed photograph technology and image recognition technology in the inspection of the track components in urban rail transit.


  • English

Media Info

  • Media Type: Web
  • Features: References;
  • Pagination: pp 373-380
  • Monograph Title: Challenges and Advances in Sustainable Transportation Systems

Subject/Index Terms

Filing Info

  • Accession Number: 01529108
  • Record Type: Publication
  • ISBN: 9780784413364
  • Files: TRIS, ASCE
  • Created Date: May 15 2014 3:02PM