EUVIMEDEUVIMED entdecken
Crossref

Automated individual piglet detection in weaning stage using deep learning

Erika Fernanda Lozano Cruz, Ilda de Fátima Ferreira, Robinson Osorio, Jairo Alexander Osorio, Verónica González, Ana Paola Montoya

Journal of Animal Behaviour and Biometeorology · 01.06.2026 · 14 (2) · 2026018

DOI: 10.31893/jabb.2026018 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Animal Behaviour and Biometeorology, online erschienen am 01.06.2026. Verfasst von Erika Fernanda Lozano Cruz und 5 weiteren Autor:innen. Die Arbeit nennt 36 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Individual recognition of animals on modern farms is critical for their management. Identification of pigs is the first step to individualizing the information of each animal for decision-makers. Therefore, the objective of this study was the detection of pigs in the weaning phase employing deep learning techniques. The weaning phase was selected due to issues related to piglet’s adaptation after nursing, presenting social, environmental, physiological and management transitions, challenges, where the piglets tend to cluster together, representing an image processing problem. The YOLOv7 algorithm and VGG Image Annotator were used to detect individual pigs in two pens separated by gender. Frames taken from videos recorded in production conditions of a small farm, in tropical conditions in Colombia, were used for training purposes. Animal postures, clustering and lighting conditions were considered. The Precision, Recall and mAP were used to quantify model performance. A precision of 98.3% was reached with a threshold of 0.9 and a Recall of 98.5% was achieved with a threshold of 0.85. Data augmentation techniques were employed to increase the dataset. The trained algorithm is suitable to detect weaning pig bodies under different conditions, being the first step for implementing precision farming techniques in this pig phase.

Zitieren

APA

Cruz, E. F. L., Ferreira, I. D. F., Osorio, R., Osorio, J. A., González, V., & Montoya, A. P. (2026). Automated individual piglet detection in weaning stage using deep learning. Journal of Animal Behaviour and Biometeorology, 14(2), 2026018. https://doi.org/10.31893/jabb.2026018

Vancouver

Cruz EFL, Ferreira IDF, Osorio R, Osorio JA, González V, Montoya AP. Automated individual piglet detection in weaning stage using deep learning. Journal of Animal Behaviour and Biometeorology. 2026;14(2):2026018. doi:10.31893/jabb.2026018

BibTeX
@article{Cruz2026,
  author = {Cruz, Erika Fernanda Lozano and Ferreira, Ilda de Fátima and Osorio, Robinson and Osorio, Jairo Alexander and González, Verónica and Montoya, Ana Paola},
  title = {Automated individual piglet detection in weaning stage using deep learning},
  journal = {Journal of Animal Behaviour and Biometeorology},
  year = {2026},
  volume = {14},
  number = {2},
  pages = {2026018},
  doi = {10.31893/jabb.2026018}
}
Kostenlos starten →