Switch blade monitoring based on image processing

Increasing reliability and safety of railway systems are of the most important goals of railway infrastructure managers which requires implementing more precise automated systems to detect defections. 90 percent of railway accidents in the Tehran railway occurs on switches, while 40 percent of which is due to careless maneuvering, 25 percent because of wrong track, 25 percent for switch defects and 10 percent for diesel car collisions in turnouts. Therefore, performance of switches is of great importance in railway tracks. Having made appropriate connections and proper track shifts, the system reliability can be substantially increased and incidents will be significantly decreased. This study consists of two main parts. Firstly, a switches and their importance in railway is studied. Secondly, switch blade monitoring using image processing is discussed. For this purpose, high resolution pictures are taken from switches with high quality cameras in order to detect defects. The algorithms used in this paper are based on clarifying the changes of blades for shifting the directions and allows for assuring switch locking.