Nighttime Vehicle Detection Using Deformable Parts Model

Abstract

Vehicle detection at night time is of great importance for applications toward advanced driver assistance system. In this paper, we propose a method using deformable parts model for night time vehicle detection. Before detection, we use Nakagami distribution to find the regions of saliency. After that, we consider the regions in which pairs of regions of saliency are almost at the same horizontal line as our regions of interest. Within those regions of interest, we apply the pre-trained deformable parts model to detect vehicles. The experimental result are provided to demonstrate the effectiveness of our method.