Safe Driving Assistant Using Head Pose and Eye Gaze Detection

Abstract—Road accidents caused by driver fatigue and
distraction remain a major concern in transportation
safety worldwide. Prolonged driving hours, lack of rest,
and inattentive behaviour significantly impair a driver’s
alertness, reaction time, and decision-making ability, often
leading to severe accidents. Traditional vehicle safety
mechanisms primarily focus on minimizing post-accident
damage rather than preventing accidents caused by human
factors. Therefore, continuous monitoring of driver
behaviour and early detection of unsafe conditions are
essential.
This paper presents a Safe Driving Assistant system that
detects driver drowsiness and distraction using Eye Aspect
Ratio (EAR) and head pose estimation techniques based on
real-time facial landmark analysis. The system uses
OpenCV and Media pipe to extract facial features from a
live camera feed. Drowsiness is identified by analyzing
prolonged eye closure, while distraction is detected through
abnormal head orientation and gaze direction. Upon
detecting unsafe behavior, the system generates visual,
audio, and vibration alerts and communicates with a
mobile application to send alert messages along with the
driver’s live location to emergency contacts. Experimental
results show that the proposed system effectively detects
early signs of fatigue and distraction, thereby enhancing
road safety and supporting intelligent transportation
systems.