Design and implementation of an adaptive headlight system model for enhanced night-time driving safety

Abstract

Driving at night poses a serious risk due to the momentary blindness brought on by oncoming traffic’s headlights. Traditional headlight designs impair vision temporarily, which increases the risk of accidents. Manual headlight settings and inadequate lighting make the issue worse. An in-depth review of the various sensors-based headlight control systems used by various automaker firms has been conducted. This study discusses a system for detecting vehicles that makes use of a camera and a regression-based machine-learning model. Using a grid formation technique on video that the on-board camera has processed, the position of the approaching vehicle is determined. A signal is then sent to a reflector system that has been specially designed to enable precise beam control with the help of a control unit made up of several electronic components, such as an Arduino UNO, a relay, and a battery. The proposed technology can increase nighttime driving safety by eliminating the Troxler effect on drivers of opposing vehicles while retaining adequate road illumination for the driver on-board.