Body Condition Estimation on Cows from 3D Images Using Convolutional Neural Networks

PABLO MANGUDO; JUAN MANUEL RODRIGUEZ; CRISTIAN MATEOS; MAURICIO ARROQUI; DANIEL JATIP; CLAUDIO MACHADO; JUAN RODRIGUEZ ALVAREZ; JUAN TOLOZA; ALEJANDRO ZUNINO

Montevideo

Conferencia; 1st International Conference on Agro Big Data and Decision Support Systems in Agriculture (BigDSSAgro 2017); 2017

Red Iberoamericana de Agro-Bigdata

BCS ("Body Condition Scoring") is a method to estimate body fat reserves and accumulated energy balance of cows. The BCS score heavily influences milk production, reproduction, and health of cows. Therefore, it is important to monitor BCS to achieve better animal performance. In practice, this time-consuming task is performed visually by expert scorers, thus given scores could vary between them due to inherent subjectivity. For this reason, several studies have tried to automate BCS of dairy cows by applying image analysis and machine learning techniques. This extended abstract analyzes these studies and proposes to use Convolutional Neural Networks (CNN), a machine learning technique from the area of Deep Learning that is highly effective in image classification and has not been applied in the literature yet. A CNN method based could improve overall automatic BCS estimations and could enable extending its use beyond dairy production.