In this paper, a sensing framework suited to intelligent vehicles design for automotive safety is presented. The extraction of the intrinsic parameters of the vehicle collected through the Controller Area Network serial bus (CANbus) is discussed and the implementation of a user interface for data visualization is shown. Algorithms for road lanes detection, vehicles detection and tracking and facial traits detection and tracking have been applied to acquired video streams and combined with CANbus data as speed, longitudinal acceleration and steering angle. Preliminary results concerning joint video-CANbus context analysis are presented demonstrating the potential of the proposed system to impact design strategies of next generation intelligent vehicles.
Smart sensing framework for automotive safety based on a dual-camera system and CANbus data
MARCENARO, LUCIO;REGAZZONI, CARLO
2011-01-01
Abstract
In this paper, a sensing framework suited to intelligent vehicles design for automotive safety is presented. The extraction of the intrinsic parameters of the vehicle collected through the Controller Area Network serial bus (CANbus) is discussed and the implementation of a user interface for data visualization is shown. Algorithms for road lanes detection, vehicles detection and tracking and facial traits detection and tracking have been applied to acquired video streams and combined with CANbus data as speed, longitudinal acceleration and steering angle. Preliminary results concerning joint video-CANbus context analysis are presented demonstrating the potential of the proposed system to impact design strategies of next generation intelligent vehicles.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.