An accident is an incident, occurring suddenly, unexpectedly, and inadvertently under unforeseen circumstances. Each day thousands of road accident victims succumb to injuries and death globally. According to studies, almost one-quarter of all serious motorway accidents are due to drivers handling the vehicle in a state of heavy drowsiness indicating that driver drowsiness tends to cause a greater number of accidents than driving under influence. Long haul drivers who put off taking a break while moving for long-distance often run a high risk of becoming sleepy behind the wheel and fail to identify this early on. By making use of the technologies of computer vision, an interdisciplinary science that deals with how computers can be made more intelligent by developing techniques that help them gather high-level understanding from digital images or videos, a solution can be developed for this problem of detecting and notifying of driver drowsiness. The main idea behind this project is to develop a computer vision system, using OpenCV, which is non- intrusive that can automatically detect driver drowsiness in a real-time video stream and will alert the driver by playing an alarm if they appear to be drowsy. This system would make use of an algorithm that detects the eye landmark points of the driver and based on this can determine if the driver is in a drowsy state or not, and appropriately sound an alarm.