Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Knowledge based anomaly detection for ground moving targets

: Katsilieris, F.; Charlish, A.


Institute of Electrical and Electronics Engineers -IEEE-:
IEEE Radar Conference 2018, RadarConf 2018 : 23-27 April 2018, Oklahoma City, USA
Piscataway, NJ: IEEE, 2018
ISBN: 978-1-5386-4168-2 (print)
ISBN: 978-1-5386-4167-5 (electronic)
ISBN: 978-1-5386-4165-1 (CD-ROM)
ISBN: 978-1-5386-4166-8 (USB)
Radar Conference (RadarConf) <2018, Oklahoma City/Okla.>
Conference Paper
Fraunhofer FKIE ()

Ground moving target indicator (GMTI) radars are commonly used on airborne platforms in order to detect and track ground targets, such as cars. An operator can use the provided tracks in order to make inference about the behaviour of targets. Prior knowledge, such as road network information, plays a key role in obtaining accurate tracks and thus helping an operator make better inference. This work shows that prior knowledge can also be used for automated anomaly detection when tracking ground targets, thus lowering the workload of an operator. Models for several target behaviours of interest, such as stop-go motion, are presented and tested in simulated examples.