DocumentCode
715807
Title
A new approach for a safe car assistance system
Author
Ben Dkhil, Mejdi ; Neji, Mohamed ; Wali, Ali ; Alimi, Adel M.
Author_Institution
Res. Groups in Intell. Machines, Univ. of Sfax, Sfax, Tunisia
fYear
2015
fDate
20-22 May 2015
Firstpage
217
Lastpage
222
Abstract
Drowsiness, which is the state when drivers do not have scheduled breaks while traveling long distances, is the main reason behind serious motorway accidents. Accordingly, experts claim that drowsy state is hard to be recognized early enough to prevent serious accidents that may lead even to road deaths. In this work, we propose a new drowsiness state detection system based on physiological signals and eye blinking. An experiment has been directed to justify the utility of the proposed approach. This system uses a smart video camera that takes drivers faces images and supervises the eye blink (open and close); also, it uses the Emotiv EPOC headset to acquire the electroencephalogram (EEG) signals. Eye detection is done by Viola and Jones technique, EEG. Finally, we have chosen the fuzzy logic techniques to classify the EEG signals and eye blinking detection to analyze the results.
Keywords
driver information systems; electroencephalography; eye; face recognition; fuzzy logic; gaze tracking; neurophysiology; road accidents; road safety; video cameras; EEG signals; Emotiv EPOC headset; Viola and Jones technique; closed eye detection; driver face images; drowsiness state detection system; electroencephalogram signals; eye blinking; fuzzy logic techniques; long-distance travel; motorway accidents; open eye detection; physiological signals; road accidents; road deaths; safe car assistance system; smart video camera; Accidents; Electrodes; Fuzzy logic; Process control; Reflection; Sleep; Vehicles; EEG; Viola and Jones; drowsiness; eye blinking; fuzzy logic;
fLanguage
English
Publisher
ieee
Conference_Titel
Advanced Logistics and Transport (ICALT), 2015 4th International Conference on
Conference_Location
Valenciennes
Type
conf
DOI
10.1109/ICAdLT.2015.7136627
Filename
7136627
Link To Document