Table 1 Overview of driver drowsiness detection measurement types.
Measurement type | Definition | Examples | Advantages | Limitations |
|---|---|---|---|---|
Biological measurements | Use physiological signals to assess driver’s internal state and fatigue levels | Brain Signal, Respiratory Signal, Heart Signal, Skin Signal, eye signal | High accuracy in detecting true drowsiness; real-time monitoring of body functions | Intrusive; requires wearable sensors; may cause discomfort or distraction |
Image-/video-based measurements | Analyze driver’s facial features and head posture via camera input | Eye, Mouth, Head, Hybrid (Eye, Mouth, Head) | Non-intrusive; cost-effective; compatible with deep learning and computer vision methods | Sensitive to lighting, occlusion, and sunglasses; high computational cost |
Vehicle-based measurements | Derive behavioral data from vehicle dynamics and driver interaction | Steering Wheel, Lane Deviation | Non-intrusive; utilizes existing vehicle systems; no need for driver contact | Indirect measurement; affected by road conditions and driving style |
Hybrid measurements | Combine two or more types to increase detection accuracy | Vehicle & Image, Biological & Image, Vehicle & Biological, Vehicle, Image & biological | Robust and reliable; complements strengths of individual methods | Complex system integration; increased cost and processing requirements |
