Building a Railroad Model using an Edge Detection Algorithm

Driver assisted and autonomous driving systems are now moving from feature to model detection using machine learning for detecting rail-road boundaries. It is ideal that these models are extracted from real world environments rather than synthetically generated models. Where possible, railroad modelling systems should be used in real-time applications that do not require much processing power. A railroad modelling systems can be used as a building block for other systems like object detection; the model serves as the region of interest to search for impending obstacles. Since the railroad modelling systems can be used for real-time and offline applications it is imperative that the railroad model be traced out as closely as possible. Most railroad modelling algorithms approximate the railroad boundaries as a linear model though this is not always the case.