Driver Head Pose and View Estimation with Single Omnidirectional Video Stream

Driver distraction is an important issue in developing new generation of telematic systems. Our research is focused on development of novel machine vision systems, which can provide better understanding of the state of the driver and driving conditions. In this paper we discuss in detail on the development of a system which allows simultaneous capture of the driver's head pose and driving view. The system utilizes a full 360 degree panoramic field of view using a single video stream. The integrated machine vision system includes modules of perspective transformation, feature extraction, head detection, head pose estimation, and driving view synthesis. The paper presents a multi-state statistical decision models with Kalman filtering based tracking for head pose detection and face orientation estimation. The basic feasibility and robustness of the approach is demonstrated with the help of a series of systematic experimental studies.