Petervari, RenéRenéPetervariGiovanneschi, FabioFabioGiovanneschiJohannes, WinfriedWinfriedJohannesGonzález-Huici, MaríaMaríaGonzález-Huici2022-03-142022-03-142019https://publica.fraunhofer.de/handle/publica/410335In the field of non-cooperative target classification micro-Doppler signatures provide valuable information on a diverse set of targets. In the recent years numerous applications were proposed in areas like road safety or infrastructure protection. However, real-world environments are often challenging and put high demands on hardware and software likewise. Often, time, memory or the signal-to-clutter ratio turn out to be critical factors. In consideration of these circumstances a computationally efficient and memory saving detection, tracking and classification system for real-world FMCW radar data is proposed and verified experimentally.en621conference paper