REAL TIME DRIVER ALERTNESS MONITORING USING MACHINE LEARNING

B.GNANESHWAR; B.VEMALATHA; G.SINDHU; GOLLA SAITEJA; P. SUMITHABHASHINI · 2026 · Crossref

DOI: 10.70849/ijsci03022614567

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the critical safety issue of driver fatigue, a leading cause of road accidents during long-distance and nighttime driving. The authors propose a real-time Driver Drowsiness Detection and Alert System designed to mitigate this risk using computer vision and machine learning techniques. The motivation stems from the limitations of earlier vehicle-based metrics, which often failed to detect fatigue early, and the high computational costs of deep learning models, which hinder deployment on low-cost hardware. The system aims to provide a non-intrusive, cost-effective solution that combines immediate driver alerts with remote notification capabilities. The system is implemented in Python, utilizing OpenCV for real-time video capture and Media-Pipe Face Mesh for precise facial landmark detection. The core detection logic relies on the Eye Aspect Ratio (EAR), a metric calculated from eye landmarks to determine eye openness. NumPy is employed for efficient mathematical processing of these landmarks. The system continuously monitors the driver via a standard webcam; if the EAR falls below a defined threshold for a sustained period, indicating prolonged eye closure rather than natural blinking, the driver is classified as drowsy. Upon detection, the system triggers a multi-layered response: an audible alarm alerts the driver immediately, while a background thread captures a timestamped image of the driver. Simultaneously, the system retrieves the approximate geographic location via IP-based geolocation and sends an automated email containing the photo, time, and location link to a predefined recipient. Testing results demonstrate that the system performs reliably under normal lighting conditions using standard hardware. The EAR-based approach successfully distinguished between normal blinking and fatigue-related eye closure, with time-based validation effectively preventing false alerts. Facial landmark tracking remained stable despite minor head movements, ensuring robustness. The alert mechanisms functioned as intended, with the audible alarm providing instant feedback and the communication module reliably delivering visual evidence and location data to remote recipients. The system maintained a stable frame rate throughout operation, confirming its suitability for real-time monitoring without specialized sensors or high-end computational resources. The significance of this work lies in its demonstration that effective driver monitoring can be achieved using accessible, open-source tools and ordinary hardware. By integrating vision-based detection with automated communication, the system bridges the gap between academic models and practical safety applications. It offers a scalable solution suitable for both personal vehicles and fleet management, providing immediate intervention and remote oversight. The authors conclude that this approach enhances road safety by leveraging simple automation and computer vision, with potential for future enhancements including deep learning integration, mobile deployment, and adaptive thresholding to further improve accuracy and usability.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• drowsiness detection algorithms
• distraction detection algorithms
• dms validation
• gaze based attention detection

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data
• Methodological Resource: tool software