KIT | KIT-Bibliothek | Impressum | Datenschutz

4D Radar Object Localization and Classification

Schönfeld, Leon Rafael; Wunderle, Yannick; Hohmann, Sören 1
1 Institut für Regelungs- und Steuerungssysteme (IRS), Karlsruher Institut für Technologie (KIT)

Abstract:

Autonomous mobile robots require functional safety perception systems, such as 2D laser scanners, which perform reliably indoors, but degrade outdoors, where snow, rain, and airborne debris frequently trigger false stops. Sensor fusion with optical modalities (e.g., RGB images) is prone to performance drops in adverse weather and lighting conditions. To overcome these limitations, we propose a radar-only classification and localization approach that transforms 4D radar ADC data into Cartesian images with object height, intensity, and Doppler velocity channels. Unlike many high-resolution radar setups, this work intentionally uses a low angular resolution sensor to reflect embedded and cost constraints of practical AMR systems. With YOLOv8, we successfully detect and classify people and bicycles in unstructured outdoor environments based on these images. We achieve an AP$_{30}$ of 0.53 for people on the SEW multimodal AMR dataset and demonstrate that multi-object detection is feasible with low angular resolution radar alone, even under challenging environmental conditions.


Download
Originalveröffentlichung
DOI: 10.1109/ETFA68377.2026.11725282
Zugehörige Institution(en) am KIT Institut für Regelungs- und Steuerungssysteme (IRS)
Publikationstyp Proceedingsbeitrag
Publikationsmonat/-jahr 09.2026
Sprache Englisch
Identifikator ISBN: 979-8-3315-7888-6
ISSN: 1946-0740
KITopen-ID: 1000199366
Erschienen in 2026 IEEE 31st International Conference on Emerging Technologies and Factory Automation (ETFA)
Veranstaltung 31st IEEE International Conference on Emerging Technologies and Factory Automation (ETFA 2026), Västerås, Schweden, 08.09.2026 – 11.09.2026
Verlag Institute of Electrical and Electronics Engineers (IEEE)
Seiten 1–7
Externe Relationen Siehe auch
Schlagwörter AMR, Object Classification, Radar, Safety
Nachgewiesen in Scopus
OpenAlex
KIT – Die Universität in der Helmholtz-Gemeinschaft
KITopen Landing Page