Behavior Analysis of Car Drivers Based on Computer Vision
This project builds an integrated advanced driver assistance system (ADAS) powered by computer vision.
It combines YOLOv8 for exterior object detection, Optical Flow for motion tracking,
and an SSD-based face detector for monitoring the driver’s head pose inside the vehicle.
A built-in Voice Assistant provides real-time alerts about hazards, surrounding traffic, and driver inattention, enhancing overall road safety.
A built-in Voice Assistant provides real-time alerts about hazards, surrounding traffic, and driver inattention, enhancing overall road safety.
System Components
Head Direction Detection
Uses SSD-based face detection to estimate the driver’s head direction.
Object Detection and Tracking
YOLOv8 detects external objects while Optical Flow tracks their motion over time.
Examples of the Running System
Wake Up! 😴➡️⚠️
Detects drowsiness and alerts the driver.
Traffic Light 🚦
Identifies traffic lights ahead and warns the driver.
Pedestrian 🚶♂️
Detects nearby pedestrians for safer driving.
Look Left 👈
Warns about hazards approaching from the left side.
System Architecture
