Design and Implementation of a Driver Drowsiness Detection System - A Practical Approach

Čolić, Aleksandar; Marques, Oge; Furht, Borko · 2014 · Unknown

DOI: 10.5220/0005012302410247

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 presents the design and implementation of a driver drowsiness detection system motivated by alarming road safety statistics, specifically that approximately 6% of accidents are caused by drowsy driving, often resulting in fatalities. The authors adopt a behavioral approach using visual input to detect signs of sleepiness, such as eye closure and head nodding. The system is designed to be algorithmically simple, computationally non-intensive, and portable to mobile devices like Android smartphones mounted on vehicle dashboards. Key requirements include robustness against lighting variations, camera motion, and changes in driver appearance, while minimizing both false positives (which annoy users) and false negatives (which are dangerous). The system architecture consists of four stages: detection, tracking, warning, and alert. It combines off-the-shelf components with novel adaptations. Face and eye detection utilize the Viola-Jones algorithm, enhanced by a user-specific skin color model based on red chroma histograms to improve reliability. Eye state classification (open vs. closed) employs a Support Vector Machine (SVM) trained on samples collected during an initialization phase where the driver demonstrates open and closed eyes. To detect nodding, the system uses a dynamic two-threshold method monitoring head position relative to the eyes, analyzing velocity and trajectory to distinguish sleepy nods from other head movements. Preliminary experiments were conducted in MATLAB to test the system's limitations and performance. A camera rotation test showed reliable eye detection within a 25-degree viewing angle from the center, suitable for dashboard mounting. A head rotation test revealed reliable detection within 35 degrees of forward gaze, though false positives occurred at extreme angles (90 degrees) due to occlusion. A real-world consistency test with natural driver movements showed high stability with minimal false positives. Finally, an SVM classification test using 571 frames from a single subject achieved a 96% success rate in distinguishing open from closed eyes. The authors note that while the sample size was small, the high accuracy supports the viability of user-specific models. The study concludes that combining standard algorithms with user-specific customization creates a robust, efficient, and accurate drowsiness detection system. By updating baseline models throughout the system's lifecycle, the approach simplifies implementation while enhancing resilience to errors. The findings suggest that this practical, vision-based method is a viable solution for real-time driver assistance systems, offering a balance between computational efficiency and detection accuracy.

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. Discovered via author_sweep_intake on 2026-05-28.

StageOutcomeToolModelPromptAttemptsCompleted
discover success author_sweep 2 2026-05-28
archive success canonical_url 17 2026-06-09
extract success cached 2 2026-06-09
clean success clean 1 2026-06-04
chunk success chunk 1 2026-06-04
embed success embed Qwen/Qwen3-Embedding-8B 1 2026-06-04
enrich success 1 2026-05-28
promote success 1 2026-06-04
summarize success llm qwen3.6-27b-prismaquant summ-v5 1 2026-06-09
tag success vector_similarity 15 2026-06-11
verify partial 1 2026-06-09

Summary generated by qwen3.6-27b-prismaquant on 2026-06-09; verification: verified_with_issues.

Topics

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

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).