A Visual Inspection System for Accurate Positioning of Railway Fastener

Jianwei LIU  Hongli LIU  Xuefeng NI  Ziji MA  Chao WANG  Xun SHAO  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E103-D   No.10   pp.2208-2215
Publication Date: 2020/10/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2020EDP7097
Type of Manuscript: PAPER
Category: Image Recognition, Computer Vision
Keyword: 
fastener,  accurate positioning,  region extraction,  visual inspection,  binarization,  

Full Text: PDF(4MB)>>
Buy this Article




Summary: 
Automatic disassembly of railway fasteners is of great significance for improving the efficiency of replacing rails. The accurate positioning of fastener is the key factor to realize automatic disassembling. However, most of the existing literature mainly focuses on fastener region positioning and the literature on accurate positioning of fasteners is scarce. Therefore, this paper constructed a visual inspection system for accurate positioning of fastener (VISP). At first, VISP acquires railway image by image acquisition subsystem, and then the subimage of fastener can be obtained by coarse-to-fine method. Subsequently, the accurate positioning of fasteners can be completed by three steps, including contrast enhancement, binarization and spike region extraction. The validity and robustness of the VISP were verified by vast experiments. The results show that VISP has competitive performance for accurate positioning of fasteners. The single positioning time is about 260ms, and the average positioning accuracy is above 90%. Thus, it is with theoretical interest and potential industrial application.