Classification and tracking of large vehicles for night driving

Night driving is dangerous already, but when driving near a large vehicle, the danger multiplies. Because the attention of the driver at night could be distracted thus fail to pay attention to dangerously large vehicles, we propose a component for ADAS systems that will help alert the drivers to the presence of large vehicles. This component utilizes a combination of rear lights detector, LBP-based Adaboost classifier and a BOF classifier that uses patent-free feature extractors/descriptors to extract and verify the existence of large vehicles. The preliminary experimental results show that this combination of setup is feasible in achieving desired results.