Identification of real-time driver distraction using optimal subBand detection powered by Wavelet Packet Transform

Many of the fatalities involved on-road accidents are associated with driver distraction. In order to reduce the possible chances of road disasters, it is essential to characterize the pre-requisites of driver distraction. While driving, the driver might get distracted by several ways such as talking on the cell phone, texting, and having a conversation with the passenger. There has been extensive research conducted to estimate driver states in recent years particularly on camera and EEG-based systems. However, camera-based systems face challenges such as privacy or latency in detection. On the other hand, Electroencephalography (EEG) based detection can accomplish more reliable detection. However, this technology requires an intrusive implementation. Electrocardiogram (ECG) is a reliable signal which can characterize the physiological changes consistently, with minimal intrusiveness, and at low cost. In this paper, we propose an ECG signal processing recipe with the aim of predicting driver distraction in real-time. Six drivers actively participated in the naturalistic driving experiment where distraction was induced by: 1) making a phone call and 2) having an active conversation between the driver and the passenger. We present an effective frequency subBand analysis using Wavelet Packet Transform (WPT). Due to high dimensionality of the original WPT features, we then applied Principle Component Analysis (PCA) for feature space dimensionality reduction. Based on our experimental results, WPT features demonstrated high information content and provided a significant statistical difference between normal vs. distracted driving scenarios.