More than 1.3 million people are killed in traffic accidents annually. Road traffic accidents are mostly caused by human error. Therefore, an accurate driving fatigue detection system is required for drivers. Most driving fatigue detection studies concentrated on improving feature engineering and classification methods. We propose a novel driving fatigue detection framework concentrating on the development of the preprocessing, feature extraction, and classification stages to improve the classification accuracy of fatigue states. The proposed driving fatigue detection framework measures fatigue using a two-electrode ECG. The resampling method and heart rate variability analysis were used to extract features from the ECG data, and an ensemble learning model was utilized to classify fatigue states. To achieve the best model performance, 40 possible scenarios were applied: a combination of 5 resampling scenarios, 2 feature extraction scenarios, and 4 classification model scenarios. It was discovered that the combination of a resampling method with a window duration of 300 s and an overlap of 270 s, 54 extracted features, and AdaBoost yielded an optimum accuracy of 98.82% for the training dataset and 81.82% for the testing dataset. Furthermore, the preprocessing resampling method had the greatest impact on the model’s performance; it is a new approach presented in this study.

One of the WHO’s strategies to reduce road traffic injuries and fatalities is to enhance vehicle safety. Driving fatigue detection can be used to increase vehicle safety. Our previous study developed an ECG-based driving fatigue detection framework with AdaBoost, producing a high cross-validated accuracy of 98.82% and a testing accuracy of 81.82%; however, the study did not consider the driver’s cognitive state related to fatigue and redundant features in the classification model. In this paper, we propose developments in the feature extraction and feature selection phases in the driving fatigue detection framework. For feature extraction, we employ heart rate fragmentation to extract non-linear features to analyze the driver’s cognitive status. These features are combined with features obtained from heart rate variability analysis in the time, frequency, and non-linear domains. In feature selection, we employ mutual information to filter redundant features. To find the number of selected features with the best model performance, we carried out 28 combination experiments consisting of 7 possible selected features out of 58 features and 4 ensemble learnings. The results of the experiments show that the random forest algorithm with 44 selected features produced the best model performance testing accuracy of 95.45%, with cross-validated accuracy of 98.65%.