Drowsy driver detection by EEG analysis using Fast Fourier Transform

Ben Dkhil, Mejdi; Wali, Ali; Alimi, Adel M. · 2015 · Crossref

DOI: 10.1109/isda.2015.7489245

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 drowsiness, a major contributor to traffic accidents caused by declining attention and delayed danger recognition. The authors propose an automatic method to evaluate drowsiness stages by analyzing Electroencephalogram (EEG) signals. The motivation stems from the need for non-invasive, real-time monitoring systems that can detect the transition from alertness to drowsiness more effectively than vehicle-based or behavioral measures alone. The methodology utilizes an Emotiv EPOC headset with 14 electrodes positioned according to the 10-20 system to record EEG data. The signal processing pipeline involves applying the Fast Fourier Transform (FFT) to 20-second epochs of time-series data to convert signals into the frequency domain. Pass-band filtering is then used to isolate specific frequency bands, primarily Alpha (8–13 Hz) and Beta (13–30 Hz), as these are most indicative of drowsiness and alertness states. From these bands, three features are extracted: Arousal (A), calculated as the ratio of alpha to beta power in frontal lobes; Valence (V), derived from prefrontal lobe asymmetry; and Dominance (D), based on beta/alpha ratios in frontal and parietal lobes. These features serve as inputs to a Mamdani fuzzy logic classifier, which employs nine rules with membership functions (Small, Medium, Large) to determine the Drowsiness State (DS). The system was validated using eight samples from the Physionet sleep-EDF database. The experimental results demonstrate that the proposed approach successfully identifies spectral changes associated with drowsiness. Specifically, the analysis revealed an increase in Alpha and Theta power during the transition from an alert to a drowsy state, with the increase being more significant in the Alpha band. The fuzzy logic controller effectively processed the extracted Arousal, Valence, and Dominance features to classify the drowsiness level. The authors conclude that the system achieves a degree of recognition suitable for evaluating drowsiness levels, validating the efficacy of using FFT-based feature extraction combined with fuzzy logic for this application. The significance of this work lies in its contribution to Brain-Computer Interface (BCI) applications for automotive safety. By providing a robust, EEG-based detection mechanism, the study supports the development of advanced driver assistance systems. The authors suggest that future work should expand this system to monitor both internal risks (driver vigilance) and external risks (road objects), thereby creating a comprehensive safety framework. The findings confirm that EEG spectral analysis, particularly focusing on Alpha and Beta bands, is a viable and effective method for real-time drowsiness detection.

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
• drowsiness
• microsleep

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