Development of a Line-Scan Vision System for Detection of Rail Surface Defects Supported by: The Internal Competitive Research Grants of Hong Kong Polytechnic University Project Timeline: September 2011 - September 2014 |
||
Typical rail surface defects: rail cracks |
Typical rail surface defects: rail spalling |
Typical rail surface defects: rail corrugation |
Project Overview | Research Team | Publications | Related Sites |
The reliability of rail tracks is essential for safety rail transit. Recent train accidents happened in Hong Kong have reaffirmed the need for developing effective systems for rail defect detection. Traditional rail inspection systems based on ultrasonic devices experienced with problems in inspection reliability, and only digital signals can be obtained from this type of devices instead of providing images showing the actual defect situations. This proposal plans to develop an advanced line-scan vision system for detection of rail surface defects. The objectives of the proposed research are: 1) To develop a prototype line-scan vision system, which can be mounted on the train with the capability of acquiring high quality images of the rail surface, and 2) To develop innovative image analysis techniques for detecting rail surface defects (rail cracks, rail spalling, and rail corrugation) from the acquired images. The developed system and methodology can contribute to the safety of railway transportation in Hong Kong. The proposed research will facilitate the advancement of the science and technology of machine vision in rail industry. The developed hardware and software systems can also be transferred and used in a variety of other potential industry applications, for example, defect detection on road or tunnel surfaces, in Hong Kong and mainland China. |
|
Principal Investigator: Dr. Bo Wu, Assistant Professor, The Hong Kong Polytechnic University
Research Staff:
|
|
For further information, contact Dr. Bo Wu Department of Land Surveying & Geo-Informatics