EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Segment Any Plant (SAP): foundation-model segmentation for plant time-series phenotyping

Alex Abbey, Yasmine Meroz iD

Journal of Experimental Botany · 25.09.2026

DOI: 10.1093/jxb/erag467 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Experimental Botany, online erschienen am 25.09.2026. Verfasst von Alex Abbey und 1 weiteren Autor:innen. Beteiligte Einrichtung: The Center for Physics and Chemistry of Living Systems, Tel-Aviv University, Tel Aviv 6997801, Israel. Die Arbeit nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Israel Science Foundation, European Union, Lautman Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Quantitative studies of plant growth and environmental responses increasingly rely on time-series imaging, yet automated segmentation remains challenging due to continuous growth, large non-rigid morphological change, and frequent self-occlusion. Traditional image-processing pipelines and task-specific deep learning models often require extensive annotated datasets and retraining, limiting portability across species, developmental stages, and imaging conditions. Here we present SAP (Segment Any Plant), a plant-focused framework that leverages the pretrained Segment Anything Model 2 (SAM2) to enable few-shot, training-free segmentation of plant time-series imagery. SAP integrates interactive prompting, automated temporal mask propagation, and centerline extraction within a web-based interface, allowing users to move from raw images to quantitative descriptors of organ shape, growth, and dynamics without programming expertise. Across multiple systems, including Arabidopsis rosette development, root growth, sunflower gravitropism, and confocal root microscopy, SAP achieves high segmentation accuracy on macroscopic organs [mean intersection over union (IoU) 0.89–0.93] and pixel-level centerline precision from single-frame prompting, with lower but promising accuracy at the cellular scale (confocal, mean IoU 0.75), which we report as a proof of concept. By reducing the need for task-specific retraining, SAP provides a transferable framework for reproducible time-series phenotyping across diverse experimental contexts.

Schlagwörter

Stichwörter der Autor:innen

Agricultural technologyComputer visionFoundation modelsPlant imagingPlant phenotypingSAM2Time series segmentation

Autor:innen und Institutionen

  • The Center for Physics and Chemistry of Living Systems, Tel-Aviv University, Tel Aviv 6997801, Israel.
  • Department of Condensed Matter, School of Physics and Astronomy, Tel Aviv University, Tel Aviv 69978, Israel.
  • School of Plant Sciences and Food Security, Tel Aviv University, Tel Aviv 6997801, Israel.

ORCID-Profile: Yasmine Meroz

Förderung

  • Israel Science Foundation
  • European Union
  • Lautman Foundation

Zitieren

APA

Abbey, A., & Meroz, Y. (2026). Segment Any Plant (SAP): foundation-model segmentation for plant time-series phenotyping. Journal of Experimental Botany. https://doi.org/10.1093/jxb/erag467

Vancouver

Abbey A, Meroz Y. Segment Any Plant (SAP): foundation-model segmentation for plant time-series phenotyping. Journal of Experimental Botany. 2026. doi:10.1093/jxb/erag467

BibTeX
@article{Abbey2026,
  author = {Abbey, Alex and Meroz, Yasmine},
  title = {Segment Any Plant (SAP): foundation-model segmentation for plant time-series phenotyping},
  journal = {Journal of Experimental Botany},
  year = {2026},
  doi = {10.1093/jxb/erag467}
}
Kostenlos starten →