EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

MicroSlicer: Efficient Inference-Time Slicing for Small-Target Detection in Thermal UAV Search-and-Rescue Imagery

Andrei-Gabriel Manta, Costin-Emanuel Vasile, Călin Bîră

Journal of Imaging · 01.11.2026 · 12 (11) · 544

DOI: 10.3390/jimaging12110544 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Imaging, online erschienen am 01.11.2026. Verfasst von Andrei-Gabriel Manta und 2 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Union.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Thermal drones help to search for missing people in the wilderness, but flying high enough to cover wide areas shrinks each person to just a few pixels, causing standard detectors to miss them. We introduce MicroSlicer, a lightweight method that improves small-target detection by slicing each frame into overlapping tiles at inference time, with no extra training required. Tested on three detector architectures (YOLO26-nano, YOLO26-small, RT-DETR-Large) using the WiSARD thermal search-and-rescue dataset, MicroSlicer improves target recall (+11.7 to +18.9 percentage points overall and up to +22.7 percentage points on small targets) over standard full-frame inference. High-resolution scaling experiments further indicate that, while MicroSlicer yields higher small-target recall on YOLO26s and RT-DETR-L, single-pass full-frame inference at 1536 px achieves higher recall and throughput on the ultra-lightweight YOLO26n, demonstrating that the trade-off between spatial slicing and global input scaling depends on the model capacity and architecture. When benchmarked against an existing slicing framework, SAHI, with matched non-maximum suppression (NMS) and a slices-only configuration, MicroSlicer achieves similar overall detection accuracy while maintaining higher target recall across all architectures and reducing the post-processing latency via compiled C++ tensor operations. Furthermore, on Transformer detectors under default merging settings, SAHI exhibits substantial latency increases and bounding-box dilation at low confidence thresholds, whereas MicroSlicer maintains stable deduplication (0.11–0.14 ms at operational confidence and 2.1–3.6 ms under dense candidate sets at low thresholds) and consistent accuracy. We also found that full-image training combined with MicroSlicer performed best for YOLO26s and RT-DETR-L (by up to 21.0 percentage points over patch-trained models), whereas YOLO26n benefited from sliced training. While currently evaluated on a single thermal dataset using a desktop GPU workstation, these results establish MicroSlicer as a practical, training-free option for edge-assisted search-and-rescue operations, with field validation on low-power embedded hardware identified as essential future work.

Förderung

  • European Union

Zitieren

APA

Manta, A., Vasile, C., & Bîră, C. (2026). MicroSlicer: Efficient Inference-Time Slicing for Small-Target Detection in Thermal UAV Search-and-Rescue Imagery. Journal of Imaging, 12(11), 544. https://doi.org/10.3390/jimaging12110544

Vancouver

Manta A, Vasile C, Bîră C. MicroSlicer: Efficient Inference-Time Slicing for Small-Target Detection in Thermal UAV Search-and-Rescue Imagery. Journal of Imaging. 2026;12(11):544. doi:10.3390/jimaging12110544

BibTeX
@article{Manta2026,
  author = {Manta, Andrei-Gabriel and Vasile, Costin-Emanuel and Bîră, Călin},
  title = {MicroSlicer: Efficient Inference-Time Slicing for Small-Target Detection in Thermal UAV Search-and-Rescue Imagery},
  journal = {Journal of Imaging},
  year = {2026},
  volume = {12},
  number = {11},
  pages = {544},
  doi = {10.3390/jimaging12110544}
}
Kostenlos starten →