원문정보
초록
영어
Every year, many car accidents due to driver fatigue and distraction occur around the world and cause many casualties and injuries. Driver face monitoring systems is one of the main approaches for driver fatigue or distraction detection and accident prevention. Driver face monitoring systems capture the images from driver face and extract the symptoms of fatigue and distraction from eyes, mouth and head. These symptoms are usually percentage of eyelid closure over time (PERCLOS), eyelid distance, eye blink rate, blink speed, gaze direction, eye saccadic movement, yawning, head nodding and head orientation. The system estimates driver alertness based on extracted symptoms and alarms if needed. In this paper, after an introduction to driver face monitoring systems, the general structure of these systems is discussed. Then a comprehensive review on driver face monitoring systems for fatigue and distraction detection is presented.
목차
1. Introduction
2. Driver Face Monitoring System
2.1. Imaging
2.2. Hardware Platform and the Processor
2.3. Intelligent Software
2.4. Main Challenges
2.5. Evaluation Criteria
3. Face Detection
3.1. Feature-based Methods
3.2. Learning-based Methods
4. Eye Detection
4.1. Methods based on Imaging in IR Spectrum
4.2. Feature-based Methods
4.3. Other Methods
5. Detection of Other Components of the Face
5.1. Mouth Detection
5.2. Nose Detection
5.3. Salient Points Detection
6. Tracking of Face and Its Components
6.1. Search Window
6.2. Adaptive filters
6.3. Other Tracking Methods
7. Symptom Extraction Related to Fatigue and Distraction
7.1. Symptoms Related to Eye Region
7.2. Symptoms Related to Mouth Region
7.3. Symptoms Related to Head
7.4. Symptoms Related to Face
8. Fatigue and Distraction Detection
8.1. Methods based on Threshold
8.2. Knowledge-based Approaches
8.3. Methods based on Probability Theory
8.4. Statistical Methods
9. Summary
10. Conclusion and Future Works
References