Detecting Driver Fatigue With Eye Blink Behavior
URL: http://arxiv.org/abs/2407.02222v1
archive: archived pipeline: cataloged verified
Abstract
Traffic accidents, causing millions of deaths and billions of dollars in economic losses each year globally, have become a significant issue. One of the main causes of these accidents is drivers being sleepy or fatigued. Recently, various studies have focused on detecting drivers' sleep/wake states using camera-based solutions that do not require physical contact with the driver, thereby enhancing ease of use. In this study, besides the eye blink frequency, a driver adaptive eye blink behavior feature set have been evaluated to detect the fatigue status. It is observed from the results that behavior of eye blink carries useful information on fatigue detection. The developed image-based system provides a solution that can work adaptively to the physical characteristics of the drivers and their positions in the vehicle
Summary
Camera-based driver fatigue detection system that uses adaptive eye-blink behavior features in addition to blink frequency. Face landmark and head-movement tracking are used to extract per-driver baseline blink characteristics so the detector adapts to each driver's physical features and seating position. The authors argue that contactless image-based detection is more usable than contact-based physiological systems and that blink behavior, not just blink frequency, carries fatigue information.
Key finding
An adaptive eye-blink behavior feature set tuned to per-driver baselines outperforms blink-frequency-only detectors for image-based fatigue detection and accommodates variation in driver position and physical characteristics.
Methodology
Image-based pipeline with face landmark detection, head-movement tracking, and per-driver adaptive blink feature extraction. Fatigue/non-fatigue classification evaluated against a feature set including blink frequency alone and the proposed adaptive blink-behavior feature set.
Sample size: Specific participant count not extracted from sections reviewed
Quality score: 5 / 5