EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Physics-Based Synthetic Training for Automatic Target Detection in Radar Imaging

Lorena Pérez-Eijo iD, Marcos Arias iD, Jorge Martínez Fernández, Ignacio Sardinero-Meirás iD, Francisco N Pérez Fernández iD, Borja Gonzalez-Valdes iD, Yolanda Rodríguez-Vaqueiro iD, José Vázquez-Cabo iD, Jesús Grajal iD

Sensors · 30.10.2026 · 26 (21) · 6875

DOI: 10.3390/s26216875 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 30.10.2026. Verfasst von Lorena Pérez-Eijo und 8 weiteren Autor:innen. Beteiligte Einrichtung: AtlanTTic Research Center, Universidade de Vigo, 36310 Vigo, Spain. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Spanish State Research Agency.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Representative labeled radar data are difficult to acquire across the target and scene variability needed for automatic detection. Physics-based simulation offers an alternative, but its value depends on whether detectors trained on simulated responses generalize beyond training targets. This work examines that question using acquisition-aware electromagnetic simulation of a D-band multiple-input multiple-output (MIMO) radar. Complex multistatic responses are generated before image formation, and simulated and measured data undergo the same coherent reconstruction and detector-input preparation. No measured target responses are used during development. Seven convolutional neural network (CNN) detectors are compared, with YOLOX-Tiny selected based on detection performance and inference cost. Across seven geometry-disjoint folds, YOLOX-Tiny achieves mean F2=0.89 and mAP0.5=0.88 for single-object scenes, and F2=0.82 and mAP0.5=0.79 for multi-object scenes. Generalization is further examined on 14 independent jointly simulated multi-object scenes and measurements from the reference D-band radar and two additional systems. The measurements provide qualitative evidence of transfer and reveal replica-related false detections and local class ambiguities in high-resolution responses. Lower-resolution measurements retain correctly detected targets alongside additional false detections assigned to other classes. These findings highlight the need for more representative synthetic radar responses and for relating image detections to physical targets.

Schlagwörter

Stichwörter der Autor:innen

D-band radar imagingautomatic target detectionelectromagnetic simulationnear-field radarphysical opticssimulation-to-measurement transfersynthetic training data

Autor:innen und Institutionen

  • AtlanTTic Research Center, Universidade de Vigo, 36310 Vigo, Spain.
  • Information Processing and Telecommunications Center, Universidad Politécnica de Madrid, 28040 Madrid, Spain.

ORCID-Profile: Lorena Pérez-Eijo, Marcos Arias, Ignacio Sardinero-Meirás, Francisco N Pérez Fernández, Borja Gonzalez-Valdes, Yolanda Rodríguez-Vaqueiro, José Vázquez-Cabo, Jesús Grajal

Förderung

  • Spanish State Research Agency

Zitieren

APA

Pérez-Eijo, L., Arias, M., Fernández, J. M., Sardinero-Meirás, I., Pérez Fernández, F. N., Gonzalez-Valdes, B., Rodríguez-Vaqueiro, Y., Vázquez-Cabo, J., & Grajal, J. (2026). Physics-Based Synthetic Training for Automatic Target Detection in Radar Imaging. Sensors, 26(21), 6875. https://doi.org/10.3390/s26216875

Vancouver

Pérez-Eijo L, Arias M, Fernández JM, Sardinero-Meirás I, Pérez Fernández FN, Gonzalez-Valdes B, et al. Physics-Based Synthetic Training for Automatic Target Detection in Radar Imaging. Sensors. 2026;26(21):6875. doi:10.3390/s26216875

BibTeX
@article{PrezEijo2026,
  author = {Pérez-Eijo, Lorena and Arias, Marcos and Fernández, Jorge Martínez and Sardinero-Meirás, Ignacio and Pérez Fernández, Francisco N and Gonzalez-Valdes, Borja and Rodríguez-Vaqueiro, Yolanda and Vázquez-Cabo, José and Grajal, Jesús},
  title = {Physics-Based Synthetic Training for Automatic Target Detection in Radar Imaging},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {21},
  pages = {6875},
  doi = {10.3390/s26216875}
}
Kostenlos starten →