Driver fatigue detection based on head gesture and PERCLOS

Author

Jian-Feng Xie ; Mei Xie ; Wei Zhu

Author_Institution

Image Process. & Inf. Security Lab., Univ. of Electron. Sci. & Technol., Chengdu, China

fYear

2012

fDate

17-19 Dec. 2012

Firstpage

128

Lastpage

131

Abstract

Single mean is always used to detect driver fatigue, this paper propose an integrated fatigue detection system. Head gesture and eye condition are two important factors of fatigue. In this system, we use Adaboost improved with CART to detect face, and then the centroid of face is calculated to monitoring the head gesture. On the other hand, we use Active Appearance Model to detect eye´s open and close status before the PERCLOS is used to judge the degree of driver fatigue. Compared to the original system, the combination of two means make this system more reliable. The experiments results show that this method is fast and effective that can be used in real-time detection.

Keywords

eye; face recognition; gesture recognition; learning (artificial intelligence); road safety; traffic engineering computing; Adaboost; CART; PERCLOS; active appearance model; driver fatigue degree; driver fatigue detection; eye close status detection; eye condition; eye open status detection; face detection; head gesture monitoring; integrated fatigue detection system; percentage of eyelid closure; real-time detection; single mean; Abstracts; Monitoring; AAM; Adaboost; Driver Fatigue; Head Gesture; PERCLOS;

fLanguage

English

Publisher

ieee

Conference_Titel

Wavelet Active Media Technology and Information Processing (ICWAMTIP), 2012 International Conference on

Conference_Location

Chengdu

Print_ISBN

978-1-4673-1684-2

Type

conf

DOI

10.1109/ICWAMTIP.2012.6413456

Filename

6413456