Computer Vision Systems for “Context-Aware” Active Vehicle Safety and Driver Assistance

Abstract

Recent developments of information technology and mobile lifestyle have forced drivers to multitask while they drive. The in-vehicle “infotainment” technology is already taking its place in the transformation of vehicles towards more intelligent and interactive devices rather than staying as mere transportation convenience. This transformation has several advantages such as easy route navigation, real-time traffic information, and staying connected with work or people while traveling. However, it has several drawbacks concerning the impact on driver cognitive load and attention sources. Therefore, it is crucial to take advantage of state-of-the-art in-vehicle technology to produce counter-measure systems that monitor the driver status and reduce driver workload adaptively depending on the context. In recognition and analysis of the driving context together with driver status monitoring, computer vision applications supply crucial information both in the vehicle (i.e., driver head and eye tracking) and out of the vehicle (i.e., lane, pedestrian, and vehicle detection and tracking, and road sign recognition). In this chapter, we provide a broad range of computer vision applications for CA-IVS from the literature and our previous studies, and we report our current research efforts.