This research presents a comprehensive system for real-time drowsiness detection and alerting in an environment. The core functionality is based on a multi-step process that continuously integrates image capture, landmark localization, drowsiness metric computation, and adaptive alarm triggering. The initial phase involves capturing images from a camera feed, providing a continuous stream of visual data. Subsequently, a landmark localization algorithm identifies key facial landmarks, enabling precise tracking of facial features. The system lies in the calculation of Drowsiness Metric (DM), which is derived from a designed formula. This metric serves as a quantifiable indicator of the driver’s alertness level, taking into account various facial expressions. A critical decision-making step follows the metric calculation. The system evaluates whether the drowsiness metric exceeds a predefined threshold value. If the metric goes below this threshold, an alarm is triggered instantly, serving as a vital alert to the driver. This alarm includes audible and visual elements to ensure rapid response. Conversely, if the drowsiness metric remains above the threshold, the system continues to capture frames and count those frames. This research exemplifies the fusion of image processing, facial analysis, and intelligent decision-making to create a responsive and reliable drowsiness monitoring system.