In recent years, fatigue driving has emerged as a significant threat to traffic safety, underscoring the critical importance of research in fatigue detection. To address this gap, this paper proposes an innovative fatigue-driving recognition method based on emotion recognition. The method leverages the facial recognition system OpenFace to analyze facial features under driving conditions using the Facial Action Coding System. It employs a fully connected neural network model to evaluate combinations of facial action units associated with ten basic emotions for fatigue detection, thus discerning fatigue during driving. Experimental results highlight “satisfaction” and “sadness” as exhibiting the strongest predictive capability, indicating a robust correlation between these emotions and states of “fatigue driving” versus “non-fatigue driving.” This novel approach offers a fresh perspective on fatigue driving monitoring through emotion recognition, potentially paving the way for more effective detection methodologies.