| Summary: | 碩士 === 國立中興大學 === 電機工程學系所 === 100 === Vehicle penetration rate rapidly increases in the word. In recent years, car accidents caused by fatigued driving and drowsy driving are raised. Driving is spiritually exhausted types of work. There are several symptom of tired drowsy driving such as awkward movements, blurred vision, trance, unresponsive…etc.. While drowsy driving happens, it accompanies with short duration sleep and losing control of driving. Drowsy driving is usually happening during high-speed driving. If the driver doesn’t take a rest or keeps spirits up, drowsy driving might cause serious traffic accidents.
Generally, measuring the frequency of nod or close of eyes are the methods to judge the drowsiness state. However, the measuring information can only be gotten during drowsy condition. Hence, it is hard to measure drowsy information and to warn drowsy driver earlier. The thesis will use frequency domain analysis of heart rate variability (HRV) to detect if drowsiness state happens. This way is different than detect by external feature. It can early and exactly detect whether short duration sleep happens. It is a profitable way to prevent traffic accidents.
The thesis proposes a drowsiness detection system, which is implemented by Cheetah development kit (CDK) and uses graphical user interface (GUI) in PC. It can immediately display electrocardiography (ECG) waveform and power spectral density (PSD) analysis of heart rate variability. The system based on embedded system and hardware system added HW-SW Bridge that integrates easily for all system to promote system on chip (SOC) implementation in the future.
|
|---|
