EUVIMEDEUVIMED entdecken
Crossref

Employing deep learning for chum salmon (Oncorhynchus keta) age estimation through scale image analysis

Robert T. Ames, Patrick D. Barry, Matt W. Callahan, Niels Leuthold, Stephen G. Wischniowski, Jodi C. Neil, Sarah Gravel, Taylor Scott, Benjamin H. Adams, Emma I. Christman, Jordan T. Watson

Canadian Journal of Fisheries and Aquatic Sciences · 02.11.2026

DOI: 10.1139/cjfas-2026-0089 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Canadian Journal of Fisheries and Aquatic Sciences, online erschienen am 02.11.2026. Verfasst von Robert T. Ames und 10 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate age estimates of salmon produced from scale growth patterns yield demographic information essential for population assessment and fisheries management. Conventionally, salmon ages are estimated from scales by trained personnel where demand often exceeds staffing capacity and financial resources. Convolutional neural network (CNN) models are a promising approach to augment manual age estimation. We compiled an extensive scale image repository from chum salmon (Oncorhynchus keta, Salmonidae) stocks throughout the North Pacific rim, evaluated two image preprocessing algorithms—Segment Anything Model (SAM) and You Only Look Once (YOLO)—to remove or reduce background noise, and trained an ensemble of three CNN models (EfficientNetV2L, NASNetLarge, and ConvNeXtLarge, weighted by their MSE) to infer chum salmon age. Agreement with expert estimates was higher for the SAM preprocessed images (93.2%) than YOLO (91.4%) across all major age classes (ages 3–5). Both preprocessing approaches showed reduced agreement for less common age classes because of inadequate representation in the model training and validation datasets. This approach may be integrated into fisheries management workflows to accelerate the age estimation process, complement manual reading, and bolster quality control protocols.

Zitieren

APA

Ames, R. T., Barry, P. D., Callahan, M. W., Leuthold, N., Wischniowski, S. G., Neil, J. C., Gravel, S., Scott, T., Adams, B. H., Christman, E. I., & Watson, J. T. (2026). Employing deep learning for chum salmon (Oncorhynchus keta) age estimation through scale image analysis. Canadian Journal of Fisheries and Aquatic Sciences. https://doi.org/10.1139/cjfas-2026-0089

Vancouver

Ames RT, Barry PD, Callahan MW, Leuthold N, Wischniowski SG, Neil JC, et al. Employing deep learning for chum salmon (Oncorhynchus keta) age estimation through scale image analysis. Canadian Journal of Fisheries and Aquatic Sciences. 2026. doi:10.1139/cjfas-2026-0089

BibTeX
@article{Ames2026,
  author = {Ames, Robert T. and Barry, Patrick D. and Callahan, Matt W. and Leuthold, Niels and Wischniowski, Stephen G. and Neil, Jodi C. and Gravel, Sarah and Scott, Taylor and Adams, Benjamin H. and Christman, Emma I. and Watson, Jordan T.},
  title = {Employing deep learning for chum salmon (Oncorhynchus keta) age estimation through scale image analysis},
  journal = {Canadian Journal of Fisheries and Aquatic Sciences},
  year = {2026},
  doi = {10.1139/cjfas-2026-0089}
}
Kostenlos starten →