YOLO Based Automatic Poultry Monitoring System

Efficient poultry management is essential to ensure the continued availability and affordability of poultry products for consumers. Considering the crucial role of poultry farming in the global economy, this paper introduces a novel automated poultry monitoring system using the YOLO v8 model. In this work, a dataset comprising visual images was collected and utilized to predict the pathological characteristics of chickens. A total of 500 augmented images with the annotated labels as scab, swelling, paralysis and healthy chickens are taken into the account. After training, the proposed YOLOv8-based model showcased outstanding performance by achieving the top accuracy in object detection, boasting an F1 score of 0.86. The proposed system transcends passive monitoring by providing real-time data and triggering alerts via mails with triggered data and insightful images. Additionally, this proposed system emphasizes chicken welfare through proper monitoring of climate conditions such as temperature, humidity, and air quality, which significantly influence chicken health and productivity. The parameters obtained from DHT and MQ6 sensors interfaced with ESP8266 provide valuable insights for optimizing poultry farm operations and improving overall chicken health to maintain a stress-free environment, fostering increased egg production and overall flock health.