383 Harnessing the Power of Computer Vision System to Improve Management Decisions in Livestock Operations. (30th November 2020)

• Record Type: Journal Article
• Authors: Dorea, Joao R; Bresolin, Tiago; Ferreira, Rafael E P; Pereira, Luiz Gustavo R
• Is Part Of: Journal of animal science. Volume 98(2020)Supplement 4
• Page Start: 138
• Page End: 139
• Publication Date: 2020-11-30
• Subjects: computer vision -- deep learning -- livestock; Livestock -- Periodicals; Livestock; Electronic journals; Periodicals; 636.005
• DOI: 10.1093/jas/skaa278.255
• Languages: English
• ISSNs: 0021-8812

Abstract

Abstract: In livestock operations, systematically monitoring animal body weight, biometric body measurements, animal behavior, feed bunk, and other complex phenotypes is unfeasible due to labor, costs, and animal stress. Applications of computer vision are growing in importance in livestock systems due to their ability to generate real-time, non-invasive, and accurate animal-level information. Such technology has emerged as a powerful tool to predict animal identification, body weight, biometric measurements, complex behavioral traits, and feed bunk score. However, the development of a computer vision system requires sophisticated statistical and computational approaches for efficient data management and appropriate data mining, as it involves massive datasets. The objective of this talk is to provide an overview of how computer vision systems can be an effective tool to integrate animal-level information and to create predictive modeling for precise management decisions. We will discuss some of the challenges, applications, and potentials of computer vision systems in livestock, and some examples to be presented include: (1) monitoring animal growth and behavior; (2) automated feed bunk management; (3) individual animal recognition; and (4) particle size distribution in total mixed ration. The development of computer vision technologies will potentially have a major impact in the livestock industry by predicting real-time and accurate phenotypes, which, in the future, could be used to improve farm management decisions, breeding programs, and to build optimal data-driven interventions.