Skip to Main Content
The aim of this paper is to present a multiple object tracking data fusion technique, which fuses radar, image, and ego vehicle odometry. The data are fused at a high level, which leads to reliable and stable tracking results providing also additional features as width estimation and the detection of stationary objects. A ldquorealrdquo application of these algorithms is illustrated on a specific use-case: the SASPENCE system, developed inside the Integrated Project PReVENT. The principle results show that such a data-fusion technique is actually able to improve the SASPENCE performances, especially in terms of extension of operative scenarios, a basic issue for the system functionality.
Date of Conference: June 30 2008-July 3 2008