Analysis and Classification of Railway Track Surfaces based on Image Processing

Railway track monitoring, especially from a remote location is difficult task which is currently carried out manually in many countries. However, if real time monitoring is performed, it would save huge loss of materials and lives of people. In addition, the system would notify for warning and immediate attention for maintenance to prevent any future such loss because of derailment. This paper describes image processing based real time monitoring of the condition of railway tracks in remote location. The developed algorithm uses image processing to analyze, classify, categorize and estimate the surface of the railway track and the intensity of the issues like deformation, crack and gap associated with it. The proposed algorithm also estimates various details regarding the track surface issues and the same can be transmitted to the nearest railway station on regular basis. Based on the received information, the remedial actions can be initiated and undertaken to avoid catastrophic accidents that cause many lives and huge damage of material.