Anomaly Detection for Chicken Rejection using Convolutional Neural Networks (#1404)

Authors

Redondo, Carlos

Carrasco, Alberto

Abstract

In the poultry industry, early detection of diseases in poultry is crucial to prevent economic losses and ensure food safety. This article focuses on the detection of anomalies in broiler chickens using Convolutional Neural Networks (CNN). Four types of anomalies affecting chickens, namely small chickens, contaminated chickens, chickens with lameness, and asphyxiated chickens, are presented. These anomalies can lead to premature removal of chickens and economic losses in the industry. Furthermore, related works that have addressed similar issues using technologies like machine learning and computer vision are mentioned. Despite advancements, technical limitations that need to be overcome to successfully implement these solutions in the poultry industry are highlighted.