Monitoring military vehicle depots
A U-Net-based approach for vehicle counting in SAR imagery
Publication date
2026-01-20
Document type
Conference paper
Author
Hochstuhl, Sylvia
Kuny, Silvia
Hammer, Horst
Thiele, Antje
Organisational unit
Fraunhofer IOSB Ettlingen
Conference
Publisher
Universitätsbibliothek der HSU/UniBw H
Book title
Artificial Intelligence in Security and Defense : Proceedings of the workshop AI4SD
First page
36
Last page
41
Peer-reviewed
✅
Part of the university bibliography
Nein
File(s)
Language
English
Keyword
Synthetic Aperture Radar (SAR)
Object counting
Object detection
SAR simulation
CNN
U-Net
Abstract
This paper presents a CNN-based approach for the localization and counting of closely parked military vehicles in Synthetic Aperture Radar (SAR) imagery. A significant challenge in this task is the separation of small, densely positioned vehicles with partially overlapping signatures. To address this, the problem is formulated as a segmentation task, and a U-Net model is trained using point-level annotations to predict vehicle locations. To mitigate the impact of class imbalance between vehicle positions and the background, the
Tversky loss function is employed, which applies a greater penalty for missed detections of the underrepresented class. Finally, vehicle counts are derived from the predicted segmentation masks by counting connected components. To validate the approach, SAR simulation is used to generate a synthetic dataset comprising labeled SAR image data that depicts military vehicle depots from different perspectives. This dataset enables model training and allows for a comprehensive evaluation of the overall concept. The results demonstrate that the method successfully detects, separates and counts densely parked vehicles. The model trained on simulated data provides a suitable foundation for the subsequent exploitation of real image data in the future.
Tversky loss function is employed, which applies a greater penalty for missed detections of the underrepresented class. Finally, vehicle counts are derived from the predicted segmentation masks by counting connected components. To validate the approach, SAR simulation is used to generate a synthetic dataset comprising labeled SAR image data that depicts military vehicle depots from different perspectives. This dataset enables model training and allows for a comprehensive evaluation of the overall concept. The results demonstrate that the method successfully detects, separates and counts densely parked vehicles. The model trained on simulated data provides a suitable foundation for the subsequent exploitation of real image data in the future.
Version
Published version
Access right on openHSU
Open access