Nighttime Vehicle Detection at Close Range Using Vehicle Lamps Information

Vehicle detection is an important function for any advanced driver assistance systems (ADAS). Due to the restrictions in data collection, most of the detection methods presented in the past few years were based on the assumption of well-lit environments. In this paper, we propose a novel algorithm to detect vehicle in the night time by the location of vehicle lamps. The proposed algorithm calculates the intensity of diffuse light, and then uses the information to capture the locations of car lamps. The experimental results show that using our algorithm can properly detect vehicles in the night time.