Development of an Intelligent night-time obstacle detection system for vehicles

Abstract

Driving at night presents serious safety risks because of the decreased visibility, which is a contributing factor in a disproportionately high incidence of traffic accidents globally. The inability of current car safety systems to function consistently at low light levels leads to a serious weakness in current technology. The goal of this project is to develop a prototype that combines RGB and LiDAR cameras to identify obstacles in real time while driving at night. By combining the advantages of both sensors, the system uses sensor fusion techniques to provide accurate and dependable detection under a range of lighting conditions. Additionally, based on detected obstacle data, fundamental obstacle avoidance mechanisms like braking or steering adjustments will be put into place. In order to determine the system’s accuracy, dependability, and potential safety enhancements, experimental testing will be conducted under simulated low-light conditions. By improving vehicle perception and decision-making skills during nighttime navigation, this project advances intelligent transportation systems.