Driving Feature Extraction Focused on Driver's Steering Behavior in Curving

Abstract

To develop a new driving assist system that can supply effective advices to help low-skilled drivers improve driving skill, we proposed a technology to extract distinctions between high- and low-skilled drivers. There are three key contributions. The first is the introduction of wavelet transform to analyze the frequency character of driving operations. The second is a feature extraction technology based on AdaBoost. The third is a simple definition for high- and low-skilled driving. Experiments were carried out using a driving simulator on a course including several curves. Results show most features appeared at the beginning and end of curves. Specially, the character that low-skilled drivers operate hesitatingly is also detected.