EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Individual Segmentation of Intertwined Apple Trees in a Row via Prompt Engineering

Herearii Metuarea iD, François Laurens iD, Walter Guerra iD, Lidia Lozano iD, Andrea Patocchi iD, Shauny Van Hoye, Helin Dutagaci iD, Jeremy Labrosse, Pejman Rasti iD, David Rousseau iD

Sensors · 31.07.2025 · 25 (15) · 4721

DOI: 10.3390/s25154721 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 31.07.2025. Verfasst von Herearii Metuarea und 9 weiteren Autor:innen. Beteiligte Einrichtung: Laboratoire Angevin de Recherche en Ingénierie des Systèmes (LARIS), Université d'Angers, 49000 Angers, France. Hauptthemen laut MEDLINE: Malus, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 70 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Commision, European Union’s Horizon Europe Research and Innovation Programme.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Computer vision is of wide interest to perform the phenotyping of horticultural crops such as apple trees at high throughput. In orchards specially constructed for variety testing or breeding programs, computer vision tools should be able to extract phenotypical information form each tree separately. We focus on segmenting individual apple trees as the main task in this context. Segmenting individual apple trees in dense orchard rows is challenging because of the complexity of outdoor illumination and intertwined branches. Traditional methods rely on supervised learning, which requires a large amount of annotated data. In this study, we explore an alternative approach using prompt engineering with the Segment Anything Model and its variants in a zero-shot setting. Specifically, we first detect the trunk and then position a prompt (five points in a diamond shape) located above the detected trunk to feed to the Segment Anything Model. We evaluate our method on the apple REFPOP, a new large-scale European apple tree dataset and on another publicly available dataset. On these datasets, our trunk detector, which utilizes a trained YOLOv11 model, achieves a good detection rate of 97% based on the prompt located above the detected trunk, achieving a Dice score of 70% without training on the REFPOP dataset and 84% without training on the publicly available dataset.We demonstrate that our method equals or even outperforms purely supervised segmentation approaches or non-prompted foundation models. These results underscore the potential of foundational models guided by well-designed prompts as scalable and annotation-efficient solutions for plant segmentation in complex agricultural environments.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

MalusImage Processing, Computer-AssistedAlgorithms

Stichwörter der Autor:innen

datasetfoundation modelsphenotypingprompt engineeringtree segmentationzero-shot learning

Autor:innen und Institutionen

  • Laboratoire Angevin de Recherche en Ingénierie des Systèmes (LARIS), Université d'Angers, 49000 Angers, France.
  • Institut de Recherche en Horticulture et Semences, Institut national de recherche pour l'agriculture, l'alimentation et l'environnement, 49070 Beaucouzé, France.
  • Research Centre Laimburg, 39040 Auer, Italy.
  • Institut de Recerca i Tecnologia Agroalimentàries, 08140 Barcelona, Spain.
  • Agroscope, 8820 Waedenswil, Switzerland.
  • Better3fruit N.V., 3202 Rillaar, Belgium.

ORCID-Profile: Herearii Metuarea, François Laurens, Walter Guerra, Lidia Lozano, Andrea Patocchi, Helin Dutagaci, Pejman Rasti, David Rousseau

Förderung

  • European Commision
  • European Union’s Horizon Europe Research and Innovation Programme

Zitieren

APA

Metuarea, H., Laurens, F., Guerra, W., Lozano, L., Patocchi, A., Van Hoye, S., Dutagaci, H., Labrosse, J., Rasti, P., & Rousseau, D. (2025). Individual Segmentation of Intertwined Apple Trees in a Row via Prompt Engineering. Sensors, 25(15), 4721. https://doi.org/10.3390/s25154721

Vancouver

Metuarea H, Laurens F, Guerra W, Lozano L, Patocchi A, Van Hoye S, et al. Individual Segmentation of Intertwined Apple Trees in a Row via Prompt Engineering. Sensors. 2025;25(15):4721. doi:10.3390/s25154721

BibTeX
@article{Metuarea2025,
  author = {Metuarea, Herearii and Laurens, François and Guerra, Walter and Lozano, Lidia and Patocchi, Andrea and Van Hoye, Shauny and Dutagaci, Helin and Labrosse, Jeremy and Rasti, Pejman and Rousseau, David},
  title = {Individual Segmentation of Intertwined Apple Trees in a Row via Prompt Engineering},
  journal = {Sensors},
  year = {2025},
  volume = {25},
  number = {15},
  pages = {4721},
  doi = {10.3390/s25154721}
}
Kostenlos starten →