Video camera and capacitive sensor data fusion for pedestrian protection systems

In this paper a cyber-physical system is proposed to detect and classify a vehicle-pedestrian collision before the time of impact in order to trigger safety functions. A device for pedestrian protection can be activated in case of a collision with an appropriate interpretation of the current situation by noncontact sensing. A system concept consisting of an intelligent sensor fusion of a mono video sensor and a capacitive sensor is presented. The described system predicts a pedestrian collision and validates the detection with a capacitive sensor. A very precise temporal and local prediction of the pedestrian collision with the bumper of a car is possible in order to activate pedestrian protection actuators in time.