Track based multi sensor data fusion for collision mitigation

This paper presents a high level fusion approach suitable for automotive sensor networks with redundant field of views. The advantage of this method is that it ensures system modularity and allows benchmarking, as it does not permit feedbacks and loops inside the processing. The proposed algorithm deals with the issue of multidimensional assignment as the test vehicle comprises four forward looking sensors. The work described here is part of the research work in the project PReVENT/ProFusion2 where the proposed algorithm has been tested in two experimental vehicles. The paper also contains results with respect to data fusion performance for the case of collision mitigation application using several testing scenarios in real conditions.