Accidents involving automobile driving are largely attributed to human error; thus, a system capable of estimating and predicting the driver’s condition is essential for preventing such incidents. This paper presents an approach that focuses on Spontaneous Blink Rate (SBR) to detect the driver’s fatigue and arousal levels. Experimental results demonstrate a correlation between SBR and the objectively estimated fatigue level when fatigue is present. Furthermore, in the blink classification method, it was confirmed that the EARM (Eye Aspect Ratio Mapping)-based method achieves low-cost yet highly accurate blink detection, outperforming the traditional EAR (Eye Aspect Ratio)-based approach.