Smart Driver Assistant

This paper proposes a smart driver assistant system that will help drivers avoid accidents during lane departures by providing prompt and quick markings of road lanes. This paper also proposes a novel system for the automatic detection and recognition of traffic signs. It detects the blobs using MSER i.e Maximally Stable Extremal Regions which provides similar results under different lighting conditions. Recognition is based on a cascade of Convolutional Neural Networks (CNN) that were trained using histogram of oriented gradient (HOG) features.