Vision based smart driver assisting system for locomotives

Abstract

Railway transportation is a predominant transportation medium in the world. The railway transportation is one of the conventional systems that subjects to many accidents annually in most of the countries. Comparatively to other transportation systems, accident preventive methods for locomotives are not in the best use. This paper presents about a novel vision based driver assisting system for locomotives. The proposed system is capable of detecting an obstacle in the railway track. If the obstacle is within the danger zone, the driver will be notified to apply sufficient amount of brake to avoid the collision with the obstacle. The danger zone is estimated based on the current velocity of the train. Furthermore, the proposed system can be fixed into any existing locomotive without undertaking any modifications for the existing systems of the locomotive and the system is cost effective compared to the other existing methods. In order to validate the overall functionality of the system, a prototype of the system has been developed and field trails have been carried out with actual locomotives. According to the obtained experimental results, the functionality of the system has been found to be satisfactory.