Multi-sensory system for obstacle detection on railways

In the current railway systems, it is becoming more necessary to have safety elements in order to avoid accidents. One of the causes that can provoke serious accidents is the existence of obstacles on the tracks. In this work, a multi-sensory barrier - consisting of infrared (IR) and ultrasonic (US) sensors- and a vision system, is proposed in order to inform the monitoring system of the existence of obstacles. Due to the fact that any sensor has detection problems which are strongly dependent on meteorological conditions, the use of different sensors for the same task is justified so that the drawback of one sensor is compensated for by the others. The high degree of reliability needed in these environments, where the safety is fundamental, recommends the use of a multi-sensory system. Principal components analysis is applied to the data obtain from the barrier and from the vision system. The use of this technique with the barrier permits concluding if there are obstacles on the tracks; and with the vision system information about moving objects is obtained.