Deep cascaded convolutional models for cattle pose estimation

•A dataset containing 2134 images and corresponding annotations was constructed.•Three deep cascaded convolutional models was developed for cattle pose estimation.•The performance of these methods were compared by experiments. Cattle pose estimation is a key step analyzing cattle behaviors and evalu...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	Computers and electronics in agriculture 2019-09, Vol.164, p.104885, Article 104885
Hauptverfasser:	Li, Xiangyuan, Cai, Cheng, Zhang, Ruifei, Ju, Lie, He, Jinrong
Format:	Artikel
Sprache:	eng
Schlagworte:	Artificial neural networks Cattle Cattle pose estimation Color imagery Computer vision Convolutional neural networks Data augmentation Datasets Deep learning Model accuracy Pose estimation Regression analysis Regression models Robustness (mathematics)
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!