Some type of sensors represent spatial distributed objects as point targets. In this case it can happen that the origin of reflections at the object's surface are not known. Data fusion approaches have to cope with this effect to use the data efficiently. Therefore, assumptions are made about these locations in order to associate targets and update positions. In this paper, we propose a model-based approach utilizing a radar reflection model of a car that is applied to expanded objects derived from laser scanner data. Our experimental results show that fusion with radar outperforms laser-only tracking in velocity estimation and that the model-based approach gains the advantage over a 'classic' approach that considers only a single reflection center at the nearest point to the sensor.
Model-based fusion of laser scanner and radar data for target tracking
Model basierte Fusion von Laserscanner- und Radardaten zur Zielverfolgung
2010
6 Seiten, 6 Bilder, 1 Tabelle, 14 Quellen
(nicht paginiert)
Conference paper
Storage medium
English
Data fusion of infrared and radar for target tracking
IEEE | 2008
|Multi-target detection and tracking with a laser scanner
IEEE | 2004
|Study on Target Tracking Based on Vision and Radar Sensor Fusion
SAE Technical Papers | 2018
|SAE Technical Papers | 2021
|