Driver Drowsiness Detection and Alarm System Using Deep Learning

Accidents caused by drowsy drivers have become a big problem worldwide. These accidents lead to a lot of deaths and injuries every year. Despite various efforts, drowsiness remains a primary factor contributing to these accidents. This study proposes a system that analyzes facial expressions in real time to check the drowsiness of drivers and will sound an alarm when the system notices the driver is getting sleepy. The proposed system uses facial landmark detection to find important spots on the driver's face, especially around the eyes. These spots help calculate eye aspect ratio(EAR), which indicates the level of eye openness. As the eyes start to close, which means the driver is getting drowsy, thus the EAR number goes down. By using both machine learning and computer vision methods together, the system can quickly spot when drivers are getting too tired. This helps improve road safety by giving drivers warnings right away, so they can react in time and avoid accidents caused by drowsiness. By putting this system into place, it could save many lives and cut down the occurrence of road accidents globally.