Fatigue Detection System using Enhanced So and Chan Method

• Main Authors: Sih-Hao Su, 蘇思豪
• Other Authors: Chen-Hao Chang
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/18187439349413507770

碩士 === 國立中興大學 === 電機工程學系所 === 101 === Nowadays, there are more and more vehicles on the road. No matter cars, scooters, even airplanes, all of them need people manipulated. When people get tired, they may feel vision blurs, and slow reaction, and can not concentrate on driving. It will make people loss of ability to control the vehicles. Safe driving requires high level of mental concentration. Fatigue driving might cause serious traffic accidents. The common fatigue detection systems use nod or blink frequency to make judgment. However, the above methods are using the information that the driver has already entered fatigue state . Hence, it can not early warn the drivers. In this thesis, we use heart rate variability analysis (Heart Rate Variability Analysis, HRVA) in the frequency domain to detect whether the driver enters fatigue state or not. Because there is a precursor of heart rate variability. we can warn the drivers earlier by this method and it’s a good help to prevent traffic accident. To improve the “So and Chan method,” an “Enhanced So and Chan Method” is proposed in this thesis. It can effectively reduce the fail detection rate. A FPGA board is used to verify our algorithm. Then we take the RRI data into PC for power spectral density analysis to get LF / HF. The results can be used to determine the driver''s state.